 Welcome. Hello, everyone. Welcome back to our second day of this exciting workshop, developing wearable technologies to advance understanding of precision environmental health. On behalf of the Standing Committee for Emerging Science for Environmental Health Decisions, I'm just really glad that you have all been able to join us here today. I am the co-chair of the Standing Committee, and my name is Kristen Malecki. I'm division director of environmental and occupational health sciences at the University of Illinois at Chicago. Today, we have a little bit about our committee. I just wanted to, for those of you who didn't join us yesterday, I just wanted to quickly give you a little bit of oversight into what we do in emerging science for environmental health decisions. We really aim to examine scientific advances that can be used to help bridge environmental health sciences with understanding environmental impacts on human health, and we understand our world is changing rapidly. And we really need to be forward thinking about how do we take advantage of all of the amazing scientific and technological advances that are happening in order to support better protection of human health. Through this, we facilitate communication amongst government, industry, environmental groups, and the academic community, and we host several workshops throughout the year. If you're interested in finding out about what we do, please visit the National Academy of Sciences website. And importantly, I want to note that this activity is made possible through generous support from the National Institute for Environmental Health Sciences. We have an amazing committee of standing committee on the use of environmental emerging decisions or emerging science for environmental health decisions that we have a 16 member group made up of both academics, as well as practitioners at the state and local level. We also have federal government liaison group that supports the work that we do that we consult with on a regular basis to advance this work. For this workshop, we have a workshop planning committee that is listed here. Dr. Haber who we have her from the Keck School of Medicine has chaired this workshop, along with many of the other excellent speak folks here on the panel, and they've been working really hard to make this a positive experience and yesterday if that was any sign of their hard work was a was an excellent introduction to this workshop. And yesterday we really covered the possibilities of wearables how wearables can help in environmental and biomedical research and we saw several different examples of ongoing use of wearables and new research that's advancing with that. So we really want to move even further forward in exploring how wearable applications can be used in the biomedical context in disease monitoring interventions and how that intersects with environmental health sciences. And then we also really want to take a very practical look at how do we move the use of these wearables into implementation. So we'll be talking about adaptation communication and other potential barriers or strategies by which we can move all of this new technology into practical cost effective solutions for advancing environmental health sciences. So just a few quick housekeeping rules this is a public webcast and workshop to discuss and exchange data and ideas. We really want folks to be active participants there is a chat box underneath the webcast video for you. Please submit your questions there. So the comments and made an idea shared during the workshop should be attributed to the individual speakers and not their organizations unless otherwise stated. All of this should not be interpreted as the opinion of the National Academy of Sciences, Engineering and Medicine, except for these housekeeping rules. I would like to share the work that's happening on social media, and the full workshop will be recorded with a brief written summary of the workshop published within a few months. So without further ado, I want to move us towards the second day of this exciting workshop, developing wearable technologies to advance understanding of precision environmental health. And I will introduce you to Dr. Aconesano from Rice University who will be leading our first session. Hello, everyone. Welcome to our session three on exploring wearable application in other research areas such as this monitoring interventions and biomedicine. My name is Aconesano, I'm a professor at Rice University. Today we will be hearing from five speakers. Dr. Shruti Mahalinga from Harvard. Dr. Lorentian from Apple. Dr. Jess Lynn Dunn from Duke University. Dr. David Armstrong from University of South California, and Dr. Vena Mishra from North Carolina State University in this session. So first of all, I will introduce our first speakers. Dr. Mahalinga and Dr. Chen. Dr. Shruti Mahalinga is assistant professor of environmental reproductive and women's health at Harvard, TH Chen School of Public Health. She serves clinically as a physician at the Massachusetts General Hospital in the department of obstetrics and gynecology where she specialized in ovulation disorders. Reproductive endocrinology and infertility. Research seeks to understand the links between environmental and modifiable risk factors on human reproduction and gynecological diseases. Dr. Mahalinga is a creator of ovulation and menstruation health study and one of the principal investigators of the Apple women's health study. Additionally, she was awarded the 2016 Endocrine Society Early Investigator Award and Edison Family Foundation Award. Next, I will introduce Dr. Lorentian. Lorentian is a physician on the Apple team, on the health team at Apple, working across health initiatives and is passionate about its mission toward empowering users to live better and healthier lives. She continues care for patients as a faculty member at Stanford Medicine. Dr. Chen was a co-founder of the Stanford Center for Digital Health and played a large role in the implementation and loadout of telemedicine and digital health across Stanford. She completed her BA at Brown, her post baccalaureate at Harvard, and her doctor of medicine and MBA at UC Irvine. So we would like to invite Dr. Mahalinga and Dr. Chen to start their presentation. Thank you so much for the opportunity to present today. We will be discussing the Apple Women's Health Study in the context of developing wearable technologies to advance the understanding of precision environmental health. The Apple Women's Health Study is funded by Apple, and I am supported by this and other research grants. We will briefly discuss menstruation and environmental exposures, the Apple Women's Health Study and exposure and outcome assessment through digital tools. The menstrual cycle is a very complex process that involves multiple organs and organ systems. Endocrine signaling is required in order to communicate between those tissues and organ systems. Critical organs and tissues include the hypothalamus, pituitary, and ovaries, and of course the functioning of the thyroid and adrenal glands. Endocrine disruption can happen at the level of endocrine signaling, or it can happen through direct toxicity at any of these tissues. And it's important to consider exposure at critical or vulnerable time points across the lifespan. Individuals today who menstruate may be able to monitor their own menstrual cycle and ovulation using a variety of at home tools, including urinary testing for luteinizing hormone, which signals for ovulation or progesterone, the hormone, the signals that ovulation has happened. Basal body temperature, menstrual cycle tracking through either an app or through paper logging. And some may also perform blood testing for some hormones. In addition to this, there is the opportunity for using sensor-based data to facilitate understanding the cycle. I wanted to introduce the Apple Women's Health Study, which is a study that involves individuals who are 18 years or older and who have menstruated once who have an iPhone and live in the United States. This study seeks to understand the menstrual cycle and how it relates to demographic lifestyle factors and how multiple data streams including sensor-based technologies might inform our understanding of the menstrual cycle. It is an observational study and a longitudinal where individuals enroll via their phone. And the data collection and the platform involves both survey-based data through the app, as well as gives the opportunity to contribute other streams of data, including data that is housed in the health kit, including the data types that you see here. We also have the opportunity to share sensor-based data through the watch or the phone, and this facilitates contribution to research for individuals as they go about their daily lives. In our initial report of the first 10,000 participants to enroll in the study, while the watch is not required for participation, 72% of participants did report watch use in the cohort. With this, I will be transitioning the talk to Lauren Chung to discuss the potential with this platform for use of digital sensors. Thank you, Dr. Mahalangaya. The Apple Women's Health Study is a great example of how we at Apple are leveraging a variety of Apple's features and tools which are publicly available to power research. So what you'll see on this next slide is a small sample of the many various signals from Apple devices that can be used in research. Apple Watch and iPhone offer features that span across 17 different areas of health and wellness. Signals from these features such as heart rate, blood oxygen, risk temperature, and mobility metrics, just to name a few, can provide information on a wide variety of internal physiologic states. The data from these features is stored securely within the health app, which serves as a repository for both data from Apple devices, and if a user gives permission for third-party Bluetooth connected devices as well. HealthKit then enables research app to connect to this data if a user gives them permission to do so. Sensors on the iPhone and Watch can also provide important signals about the external environment, signals like environmental sound levels. Now I want to touch on some of the tools available to researchers to aid in the creation of study apps. ResearchKit and CareKit are open source frameworks that allow researchers to create apps for both medical research and care pilots. They're freely available for anyone to use and include modules that we know can be important to research, including informed consent, survey tool, and active tasks like the trailmaking test. I do want to touch on privacy for a minute because it's so paramount to everything we do as physicians and researchers. At Apple, we believe privacy is a fundamental human right. That's why all health and fitness data is end-to-end encrypted on device when passcode, touch ID, or face ID are enabled, and when stored in the cloud, it's end-to-end encrypted in transit to and on our servers. We believe that people should expect the same level of confidentiality from their devices as they do from their doctor. So we designed all of our features with this principle in mind. Users get to choose exactly what data they share, who they share it with, and can revoke that sharing at any time. And we designed our research tools with that same principle in mind. Because wearables like the Apple Watch and tools like Research Kit and Care Kit can be so powerful in enabling new discoveries, I want to share with you the Investor Data Support Program that we created. So this program grants researchers Apple Watches to enable their innovative new studies. We also have clinical and engineering teams available to provide technical feedback and background support as researchers use our open source tools. We do want to highlight one study kicking off at Texas A&M in Stanford that highlights how signals from the Apple Watch and iPhone can enable research to better understand the effects of the environment on physiology. So the study teams will be providing firefighters in California and Texas with an Apple Watch with the goal of understanding how wildfire smoke impacts heart health. They plan to monitor heart rate and rhythm, sleep, blood oxygen activity, data, and more. And then the firefighters will also wear an air quality monitor and will complete surveys related to sleep, activity, and wildfire smoke related symptoms. And that's just one of the exciting ways these devices are being used to advance research in this space. Now I'll pass it back to Dr. Mahal and Guy to close this out. Thank you so much, Dr. Chung. Bringing it back to the Apple Women's Health Study shown here is a snapshot in time with the circles representing an individual's age across the reproductive lifespan. With the potential for time varying longitudinal exposure assessment in a study like the Apple Women's Health Study. With some potential for both outcome monitoring at the level of the menstrual cycle and exposure assessment at different time windows such as that related to the menstrual cycle and phases of the cycle or more broad spectrum and longitudinal, for example, at acute reproductive transitions or even chronically over the lifespan. So with that, I wanted to thank our participants, our collaborators at the National Institute for Environmental Health Science and our sponsors and collaborative team at Apple and for your listening. Thank you so much. Thank you very much for your presentation, Dr. Mahal and Guy and Dr. Chung. And we will take all questions during our Q&A session at the end of all the presentation in this session. So next, I will introduce our next speaker, Dr. Jasmine Dan. Dr. Jasmine Dan is assistant professor of biomedical engineering and bio statistics and bioinformatics at Duke University. She includes multi-omics wearable sensor and electric self-record integration and digital biomarker discovery. Dr. Dan is the director of the Big Ideas Laboratory whose goal is to detect, treat, and prevent chronic and acute diseases through digital health innovation. We are currently a PI of the COVID Identity Study to detect and monitor COVID-19 using mobile health technologies. Dr. Dan was an NIH big data to knowledge postdoctoral at Stanford and NSF graduate research at Georgia Tech and Emory, as well as a visiting scholar at the US Centers for Disease Control and Prevention and Audio Bascular Research Institute in Madrid, Spain. Research has been internationally recognized with media coverage from the NIH director's blog to Wired, Time, and the US News and World Report. We'd like to invite Dr. Dan to start our presentation. Hi. It's lovely to be here. Thank you so much for the invitation to speak. I want to make sure that you're able to see the full screen. Yes. Okay. Fantastic. Great. So I will be talking today about work that I've been doing with digital biomarker discovery on how we can use wearables for early disease detection. So very much in line with the other talks that we've heard so far. I first want to go ahead and define specifically what I mean by a digital biomarker. So a digital biomarker is digitally collected data. So that might be location from a smartphone GPS. It could be continuous heart rate measurements from a smartwatch continuous temperature measurements and so on. And that are transformed into indicators of health outcomes. So these health outcomes could be something like a traditional clinical metric. Maybe an RT-PCR test, for example, of an infection. And essentially what we're doing here is we're mapping information that's coming from digital devices to information that would more traditionally be collected. And so what this looks like if we want to put this mathematically is that we have some target outcomes on the left-hand side here of the equation. I promise this will be the only math in this talk. And what we've got here is we're trying to essentially build a prediction algorithm or understand how close we can get to our outcomes of interest using the wearable sensor data that we have. So that could be any of the metrics that were described in the previous talk or more and bringing those together in some way to develop this function. And this function can be anything from a simple step function to linear or logistic regression all the way up to neural nets. So we can really get creative in what the design is. And what I'm going to be talking about today is some of the efforts that we've been doing to actually develop these digital biomarkers in different areas of interest. So we can imagine that on that left-hand side of the equation, our outcomes of interest are traditional clinical or contextual measurements of infection. So how contagious a person might be if they have an infection or if somebody's not infected, their susceptibility to contracting illness, some lifestyle interventions and how we might see their effects. We're very interested in circadian rhythms and circadian disruption, and I'll briefly mention some new work that we're starting on the effects of wildfires, which I think is very relevant to this audience. So I'll talk about three different case studies that are in this space. The first will be on infection detection. The second will be on pre-diabetes detection, and the third will be the study that we're kicking off on wildfires. So this work stemmed from work that I had actually begun as a postdoc at Stanford and then continued on as I joined Duke, where we have these very interesting data sets where people either have an infection as a part of sort of natural monitoring in the natural environment, or they have been a part of challenge studies here at the Duke Human Vaccine Institute. And so in either case, what we have is people who have some sort of respiratory infection. In this case, the infections that we had looked at are H3N2, H1N1, human rhinovirus, RSV, and COVID-19. And we have three different types of measurements that we would have on the left-hand side of that equation. Symptoms that might be observable like number of sneezes or coughs. The symptoms that we call unobservable like pain or nausea. And then we have tests that are more of a ground truth. And then on the other side, we have wearables. In these different studies, we've used different wearables to measure things like heart rate, temperature, accelerometry, and so on. And two of the areas that we're working on developing these predictive algorithms in are first distinguishing between whether somebody is infected or not infected. And on the other side, looking at the trajectory of illness. So if somebody is infected, what will that infection look like downstream? And one of the obvious applications here when we think about COVID-19 is if somebody shows up in a hospital and is COVID positive, can we predict whether or not they may end up needing a ventilator or some other limited resources? And so one of the application areas that we pursued pretty heavily during COVID-19 was in understanding how we could address diagnostic testing shortages at the onset of new pandemics. And so we saw, I'm sure many of us experienced firsthand at the beginning of the COVID-19 pandemic and even at points when there were peaks that it was difficult to get diagnostic tests. And the reason for this was because there was so much demand and production was not meeting demand at that time. And so what we wanted to know was could we use data coming from wearables to better inform how you allocate diagnostic tests when you are in a resource limited setting. And so what we did was we launched the study co-identify in March 2020. So it was really, you know, building the plane while flying it effort, as many of us were doing at that time. We were able to, because this was a completely remote study, consent about, it was actually about 7600 people total. And of those, the data that I'm showing here is people who tested negative for COVID and people who tested positive for COVID. And as you can see, there are clear signals from the wearable devices where we have changes in these digital biomarkers prior to somebody having a positive COVID test. And so what this indicates to us is that there are some signals that are indicative that somebody is likelier to be infected with COVID. And in this particular application, what we could use that for is to more intelligently allocate tests when we have a limited number of them using this data coming from smartwatches. So this is one of the studies. The other one that I want to briefly present is work that we're doing in cardiometabolic diseases. So the problem that we're trying to address here is another public health challenge that one in three people in the US have prediabetes and 90% of them don't know that they have it. And so what we've been doing is learning from the digital biomarker data, whether there are early indicators or even indicators that somebody may be pre-diabetic today that we could leverage to give people more information. So this is just one set of data from this much broader study that actually has a new R01 that's just gotten funded that we're excited to continue to pursue this work. And what I'm showing here is we have data coming from continuous glucose monitors, which are a slightly invasive device where you have a needle that has to be sort of entered into the skin. And what we're trying to do here is make predictions on people's glycemic health using A1C as the proxy measurements. This is the American Diabetes Association model showing the root mean squared error and the mean average percent error. The models that we built using the CGM and then using just smartwatch data and no invasive methods whatsoever are comparable, if not better than the American Diabetes Association estimated A1C model. What we can see here from this is that we could use just a smartwatch to have digital biomarkers that are indicative of somebody's A1C level, which is indicative of their glycemic health. What we hope to do with this work is on a much larger scale and deploy a technology like this for the 117 million wearables that currently exist on people's wrists to screen for pre-diabetes and alert people so that they can get the clinical test, confirm whether or not they have diabetes and start managing their disease. The last application that I'll mention relevant to this group is monitoring cardiovascular effects of wildfires. This is a new study that we're about to bring online where we will have 52 participants either using or not using in-home HEPA filtration and using wearables such that we can measure if there are cardiovascular effects to wildfire instances. The last point, I'm going to actually skip this summary and I just want to mention that there are a ton of resources here for digital biomarker discovery. So if this is an area of interest to you, I really want to encourage you to check out dbdp.org. There's a bit link here if you're interested in checking out case studies. Essentially what this resource is is data, code, algorithms and educational resources to be able to start the process of discovering digital biomarkers. So if you're interested in using this resource or contributing to it, please feel free to contact me or just log on to dbdp.org. So with that, I want to thank the many people who have contributed to this work and we'll be glad to answer questions during the end of the session. Great. Thank you, Dr. Dan for your wonderful presentation also sharing very useful resources. Next, we will have Dr. David Armstrong. Dr. David G. Armstrong is a professor of surgery with tenure at the University of Southern California. Renowned for his work in tissue repair wound the healing and diabetic foot care. He has over 600 650 peer reviewed research papers and 100 books or book chapters to his name. And is a founding co editor of the ADA clinical care of the diabetic foot. He co founded the Southwestern academic limb salvage alliance and these USC NSF funded center for stream health care in place. Here he marges consumer electronic wearables and medical devices to enhance patient care. Dr. Armstrong has received numerous upward by national and international institutions, including the Georgetown Distinguished Award for diabetic limb salvage and the 2023 ISDX current backer as a first podiatric surgeon to be named fellow of the Royal College of Surgeons Glasgow and the founding president of the American limb preservation society. He has made significant strides in his field. Dr. Armstrong is committed to ending preventable invitation within the next generation, a goal that he advances through fostering innovative interdisciplinary research, clinical practice and advocacy. We'd like to invite Dr. Armstrong to begin with presentation. Well, thank you very much, Professor Sano, and I hope you can all hear me all right. If you can't hear me I would ask you just to relax and take a nap for 10 minutes. How does that sound. But this has been absolutely spectacular. I mean, one wonderful talk after the next. I'm going to try to get you to care about something that you probably haven't thought much about, which is really the end of the body and the foot and we'll talk about that in a minute we give all of our slides away just please ping me and we ask you just to copy them and make them better. A lot of the work here is not only from the National Institutes of Health but also from the NSF, and this really awesome great five six university program called the Center to stream health care in place so. There's more on that later, you heard a little bit about this already yesterday from Bijan, and you already heard from Dr. Sano what we're going to talk about you're going to say well my gosh feet. What does this have to do with environmental health but hopefully by the end you're going to see how a lot of this stuff really can can come together. But first this you've already told you this is what I do for a living I'm, I'm a, I should say a tow mechanic and you know when you work out at the end of the body. And I think of maybe two great gifts there. One of these gifts is in this era of hubris. You know I can't think anything that's more of an expression of humility than looking after someone's feet that it's an expression that crosses borders and religions and even time. And the other great gift is that when you work on this anatomic peninsula. Instead of at the center of the body it forces you to collaborate but then have some perspective with the anatomic mainland and we're collaborating with folks doing everything from spray on and spread on skin to wearable robots and. It's just such an exciting time to be doing this and, and that's kind of what we're going to be sharing in this area but first, the perspective bit and the perspective in diabetes and lower extremity is not a good one. And right now, foot complications are common they're complicated and they're costly. They're often ignored. But hopefully for the next few minutes we'll talk about this before we talk about some of the solutions. So let's look at some of these data, just putting it into perspective here. Every second now someone around the world develops a wound on their feet they wear a hole in their foot like we wear a hole in the sock. Half of these wounds will get infected. And 20% of those infections end up in hospitalization and every 20 seconds now someone around the world has an amputation. And we'd like to think, ladies and gentlemen that that times up for this because this affects people of color. Much more than than not and this is a zip code lottery that we see around the country and around the world and I think we can make a big difference in this area. But first, a little bit more of the bad before we get into the good. It's not fair to compare one terrible thing to another but look if we compare this to cancer. If someone gets a wound, you know 30% of these folks are going to be dead in five years. If they have a minor major amputation it's up to 50 and then 60%, even up to 80% with other comorbidities. And then even after healing these wounds, there's a very high rate of recurrence in this population these folks just like with cancer are in remission. So we have a lot of opportunities to affect change in this population and, and it's really an algorithm if you will, that's kind of a, it's maybe three parts of the algorithm, there's really two variables that we can play with surgically medically mechanically, technologically, in the face of neuropathy or the loss of protective sensation people can wear a hole in their foot, like they wear a hole in the sock as I told you, and we can play with pressure. We can play with cycles of activity. And it's really an imbalance of these things that we have to try to bring back into balance. But before we do that before we address these little wounds that become big wounds, maybe we can measure what we manage a little better and we haven't done this very well in surgery and medicine. It's been positively medieval up until fairly recently and there's a lot going on now. That's very, very exciting to help us marry a little bit of engineering with a lot of the work that we're doing in the clinic and in the operating room, or if you're British, the operating theater. So let's first talk about some really cool kind of epidermal electronics and other solutions that may help us kind of help form a closed loop system. These are data from my friend way out right down the street to Caltech and where we're working on various kinds of smart managers with them and with the Institute, but you also see here, you heard from Christina Davis in that spectacular day yesterday talking about things like looking for volatiles. Well, there are a lot of other things that we can do in addition to volatiles where we can measure infection, we can measure pH change and we can actually affect change by putting in some therapeutics and creating almost a closed loop system to try to heal these wounds. But before and even as we're healing these wounds, we have to understand these patients don't have the gift of pain. So we have to be able to protect them. So how do we do that? Well, we have to do something we call offloading, which is kind of a physical characteristic, but these folks can't feel anything. So how can we get, can we get them into something that might help them? And we have a really cool tool in the study right now where we're working with the National Institutes of Health and the smart boot now that actually marries really good design that people will want to wear. This was designed by Mike DiTullo, who designed the second and third generation of Air Jordans, but now this tool can have a gyroscope right on it and actually it can tell our patients when they're wearing it, when they're not, their nurse, their doctor as they like, and maybe instead of doing something to these patients or they can do something with them and this is such an exciting time because we can help these folks kind of dose their adherence and dose their activity and really be working with them rather than do them. So now after we've healed the wound by offloading it, by monitoring it and measuring it, how do we reduce the risk of recurrence in remission? Well, this again is an engineering tool marrying that with work that we might do in the operating room or in the clinic. And we can dose activity just like we dose a drug too high and we get toxicity with the drug too low. We don't get the benefit of the drug too high with a patient with diabetes and they'll get a wound too low and they won't get the cardiometabolic benefit. The trouble with this, if we look at this from Dan Lieberman's lab in Boston, we see that the foot is just really beautifully balanced. But with our patients, activity is not balanced and we see here that it's not just the number of steps that folks take but the actual variability in that activity that we see that can proceed ulceration. But we can predict that now and plug it into an entire formula. This is a simple free formula that we can add an AI Sherpa to that can give a patient a high and a low ratio that can really affect positive change. But how can we detect that even further? You know, if you want to sound smart and who doesn't want to sound smart, I'm trying and failing but there's kind of hope for you and you're talking about any chronic disease, just say inflammation and walk away and you'll be right and we can detect inflammation beautifully here in this population just looking for looking for asymmetries and we can start using thermometer to dose activity, just like we would use give someone insulin and check that by checking their glucose. This can be done through a variety of form factors smart socks have been around for quite a while. We've been working with intelligent textiles, but what's really cool now finally. We're seeing people starting to subscribe to these sorts of beautifully designed things that used to just be a science experiment. And you see people like grandma who worked with us some time ago, actually now going to the consumer electronic show and going up against augmented reality and reality reality socks seem to be winning in this case. We're working on with Zoltan Pataki here you see this great work in Geneva on smart shoes and in souls now that can get out of their own way. And, and we are doing the same thing with with tools we can maybe even form an AI based kind of closed loop system in this population. But you know, just in the last minute or two, you know, wearables are really great. It's a whole session here but they're getting kind of passe. What about injectables. Now that's kind of cool and now there are tools and some of them are being worked on by friends and colleagues like Natalie was new ski, where we can inject tools and hit them with a laser, just a simple. Just a hydrogel looking at the analytes on that and measuring not only oxygen but also glucose and many other solutions to affect change in this population. And the other really cool things are actually entire implantables, like smart blood vessels. This is a tool now where you can pair with it on the back table in the operating room just like you pair with your Bluetooth headset, and you can look for turbulent flow. And this can be monitored, not in the clinic. It can be done entirely in the home where we can see a bypass that's about to go down before it goes down this is a PTFE conduit. There'll be other tools that could be very, very helpful here too. But bring it all together. I think the masters of the universe are going to be the women and men that can make all of these things communicate, which is why it's so exciting to hear so many brilliant folks at this symposium thinking on this. We're seeing this great work from a company called Foreside Health that has been working with our C2 ship and March scuba can put folks who are now working with entire smart homes looking not only at depth sensors and thermometer, but also motion sensors, watches, having all of those with folks in extended care facilities, allowing them to be mobile and have, you know, the hospital not only hospital free, but activity rich days, and even dignity fill days and really, at the end of the day, that's the big idea. So, Madam Chair, ladies and gentlemen, you know, often when you treat these patients you hear, you know, you're just married to them, and that term marriage is used in a negative way but hopefully with a little bit of technology and tenacity and a little humility. We can help these folks move through the world a little bit better with some enhanced perspective because I think that's what we all deserve, no matter what our health and no matter what our age, Madam Chair, everyone. Thank you so much. It's been an absolute pleasure to be here. Thank you very much. Dr. Armstrong for your wonderful future thinking presentation. Next, we will have our last speaker, Dr. Vienna. Dr. Vienna is a director of the National Science Foundation nano systems engineering research center on advanced self powered integrated sensors and technologies. While her background and training as electrical engineer is on advanced high performance silicon devices, she has spent the last 15 years of her career in integrating these technologies with non traditional technologies, both in structure and in functionality. She's a distinguished professor of electrical and computer engineering at North Carolina State University, and the 2012 I totally believe she has authors and co authors over 150 papers in the areas of state of the art, low power devices, nano scale, magnetics and energy hardest. Dr. Michelle was the recipient of the 2001 National Science Foundation president. The 2011. Distinguished engineering research. And 2007 outstanding alumni research. And 2016. Please join me. Thank you very much, Professor Sano. It is really my great pleasure to share with you all of our work that we have been conducting in the National Science Foundation assist center. This is a team of multiple universities multiple disciplines, all working together to build wearable systems to to advance personal health and personal environmental health as well. So about 12 years ago we had this vision of building wearable devices that not only measure health parameters, such as heart rate and motion and activity, but also environmental parameters. And to do this we wanted to achieve a few disruptive traits for these wearable devices we wanted them to be always on so that they would provide us with long term data. And we wanted them to also also collect different kinds of data. We wanted them to be wireless and comfortable. So they would be passively on. And we wanted to also be able to provide machine learning and algorithms on the wearable devices itself for real time actionable information. And our mission of the assist center that we, like I said, have been working in the last 12 years, needless to say, there are some key advantages of long term health monitoring, many of the wearable devices today they have to be charged regularly. The more sensors you put on them the more charging you have to do. But if there was a way to break out of that equation where you could provide continuous monitoring without the user having to change or charge the battery, then you can really open a paradigm in terms of critical positive outcomes. There are many advantages of long term health monitoring but in the context of this workshop. There is a huge opportunity to connect health and environmental toxins, especially because we are not always aware in advance what environmental toxins will be exposed to, whether it's a building or a city or a location. And so having these sensors that are continuously on collecting that environmental data, but not only environmental data at the same time and in real time, collecting your health data, and then using these algorithms to correlate the two is the opportunity that's afforded by continuous long term monitoring. And so we knew immediately that to achieve this, this goal, we have to build devices that were always on. And so our wearables do not rely on a battery charging process. We're wanting right from the beginning to harvest power from the human body use that power in the form of thermal thermal heat and thermal motion, and other sources, and make sure we're maximizing the power that was coming out from the body. And then on top of that we have to make sure that the power that we consume whether it was in the sense, sensing, or the computing or the radios to do real time communication was also as small as it could be. And then finally, it had these devices have to be comfortable and adopted by a variety of individuals, and so wearability and the data that comes out of that was also critical. So I will very briefly just show you the kinds of things that we have done to keep this equation active. And then I'll focus most of the talk on the on the environmental use case of asthma. In order to make sure that our power that we're generating from the human body is as high as possible. We have worked on several technologies in the center. I'm showing a snapshot here we have built a state of the art flexible thermal electric generators that take your body heat and convert them to voltage. We also built a state of the art piezo electric and flexo electric materials that take your body motion and convert that to usable power. We're also taking advantage of what's already available such as ambient RF and converting that to usable power using antennas that are embroidered into fabrics, and because batteries have a limited amount of recyclability or charging cycles, we're also, we also have built super capacitors, and I'll be happy to share more about this if anybody's interested on the electronic side we have really destructively lowered the power consumed components, whether it's the analog front ends of the chip whether it's the energy management or the radios, our work in the last several years have has really brought this this consumption down. What that means is, is that we have lots of power available to add different kinds of sensors on to our wearables. And in the sensor area we have been working on integrating your bioelectric sensors such as your ECG or EDA biophotonic sensors, for example 12 wavelengths of optical sensing that allows us to do much more than just PPG. We have inertial sensors we have a whole host of biochemical sensors that we have built in our center that collects sweat sweat passively without having to run on the treadmill. And then also to complete the picture we have also added environmental sensors, which will be the focus of the rest of my talk. But one more thing before I get into the into the environmental sensors. We also have spent considerable effort in integrating these systems using smart textiles using normal materials, using flexible electronic platforms to really allow a seamless integration and an adoption by by the user. So with this approach we have targeted multiple use cases in our center. The list is shown on the left here and and there's more and more research coming out that many of these use cases actually do have a connection with environmental toxins. There has been work showing that Alzheimer's is related to environmental exposures, for example, asthma is also related to and for sleep as well. So, before I get into asthma here are the different kinds of wearables that we have built. We would love to partner with anybody who wants to try out these wearables and specific studies. And here you can get all this information going on our website which I'll flash at the end as well. So I want to spend the next three and a half minutes in talking about asthma, which was one of our early use cases, and there has been a direct correlation of environmental toxins like ozone with asthma, and also with cardio degradation along with lung, lung degradation. And the metrics are and the numbers are in the literature, there are a lot of Americans who live in areas where the ozone levels are already not meeting the standards, and we wanted to build a wearable device that would provide continuous monitoring of environmental ozone environmental and volatile organic compounds, along with heart rate, respiration, cough, wheezing and activity. And because we want these systems to be continuously on we have to make sure that the power consumption was a fraction of what was available out in the commercial space. And so to this point we built two different technologies of gas sensing. One is based on ultra thin layers of metal oxides using atomic layer deposition. The ultra thin layers allows us to get very very low power consumption and along with very high sensitivity. And the second technology we built was gravimetric sensing based on MEMS technologies, specifically capacitive micro machine ultrasonic transducers. These two technologies were integrated together to provide different types of transduction mechanisms. And we integrated these into wearable systems for asthma, which I'll get to in a second. I also wanted to mention that in addition to metal oxides, the gravimetric sensors can be functionalized with different kinds of polymers that can allow us to get different volatile organic compounds and not only volatile organic compounds but specific VOCs that might be connected to different conditions, such as the ones that you see here. And so with with combining these two technologies we're able to and machine learning algorithms, we can identify variety of different VOCs and variety of different gases that are connected to asthma. And with the asthma platform we have both a health platform which is on the chest, and a risk platform which is which is one on the risk to allow gases to come into this platform. But the risk platform is also measuring several health parameters as shown in these bullets here. We tested the system out in numerous number of clinical studies with our partners at UNC Medical School, and we have got lots of very informative data showing that in first of all that are that are wearable ultra low power gas sensors are actually able to follow the gold standard ozone measurements in the in the chambers that the clinicians use. But on top of that there was also some new findings that we are uncovering where even in healthy individuals when they're exposed to very low levels of ozone, we are able to see a decrease in lung function, and also a degradation of the heart rate variability. You can find the details of this in this publication, but this suggests that if healthy individuals are suffering from very low doses amount of ozone in here, then the respiratory conditions for asthmatic individuals is expected to be a lot more diverse. And we also looked at at the environmental design the enclosure design for a system that does both both health and environmental sensing, and we found that is essential to design the system to allow air flow in the optimal manner, so that we get gas gases to the sensing element without compromising the robustness and and the degradation associated with contaminants. And so some of these details are provided down here. And where we are taking this work. Next is shown here we have two types of directions, we are looking at the volatiles coming out of the skin. And so some of the same technologies that I mentioned previously, we're now integrating them into a skin watch, and using the same sensors to now look at what's coming out of the skin under different conditions of disease, but our initial protocol has used some simple techniques for clear discern differentiation in VOC is coming out of the skin under fasting conditions, non fasting conditions and for example alcohol intake so this is another type of biochemical sensor that's based on gases. And finally, we are also building up the number of sensors that we need. So this is using a Enos array that's compatible with CMOS, and here we can build up from one or two sensors to 16 sensors and many more, where each sensor is individually addressable in terms of its operating temperature so this is work being done to further enhance the performance of the gas sensing set up. And we also have two companies that have been coming out of the work that we're doing in this sensing space, shown here. And so I think my time is up. So I'll just say that there's a huge opportunity for combining environmental sensors with wearables, and then the same gas sensing devices can also be used to measure the human volatile. And of course there are challenges along the way which makes the work even more interesting. So with that, I will stop. Thank you very much for very very exciting presentation. Now, we will be starting our Q&A session. So please everyone, please feel free to post your question using the Q&A function below the video player online. So we'd like to ask the question to Dr. Mahalinga and Dr. Chen. In the Apple Women's Study, what air quality related health endpoints were measured? How is the performance of wearable air quality sensors compared to fixed on site air quality sensors? Thank you for that question. We have not yet started measuring air quality in relation to participants in the Apple Women's Health Study, but that is an area of interest. Then next question is to Dr. Dan. Which types of wearable sensors data are utilized for comprehending COVID infection? Additionally, what kind of wearables has been employed for this purpose? Thanks for this great question. What we used in our study was really consumer wearable, so measurements of heart rate movement and sleep. We can imagine that as we can get to higher resolution data and more types of sensors, we can have higher resolution picture of perhaps what type of infection somebody might have and that's one of the directions that we're going next. Thank you. A question to Dr. Armstrong. What types of future closed loop systems do you foresee for food? Which treatment could be delivered in variable control, control doses at variable times, like ensuring it's delivered for elevated chemicals concentrations? Yeah, wonderful question. If we have, so our goal really for kind of tissue repair, but also for preventing these things is to try to identify problems and then deliver them. We think that using looking for various analytes like metalloproteases and then hitting them with an anti-metalloprotease could be a valuable option. We think we could be delivering even mechanical things, believe it or not, like electrical stimulation through some of these tools, which actually has some data to support it. Even topical oxygen now is showing surprising benefits. So things that could be inexpensive and delivered locally could be helpful on the prevention side of things. It's equally exciting and of course there we're seeing these problems rising logarithmically so there's even more of an opportunity and identifying inflammation through thermometer and dosing the activity in that fashion with a high and a low. Again, just like people dose their insulin by checking their glucose and having some of these tools that could actually get out of their own way like some of the shape memory analyze or even just physical systems. It might be really, really helpful in this population. I think there are women and men working on these sorts of things right now, but there's a lot of brain power that is still needed in this area. And it's these problems are getting more common, not less. So there's a lot of work to be done. Thank you so much for your answer. The next question is going to Dr. Mishra. Apart from Osun, which are other individual common environmental VOCs that can be measured by your sensors. What are the sensors for buying you for right. Other chemicals or are these sensors for benzene naphthalene and other petroleum related chemicals. We have built sensors that can measure a lot more volatile organic compounds than than just ozone. We have specific sensors for ethanol toluene xylene hexanol octanol and so on and so forth along with formaldehyde and acetone and so forth but also there's an opportunity that as we increase more the number of sensors in a given array to a larger and larger number, we can we can now start measuring almost anything provided we have the right machine learning algorithms in place. Trained algorithms in place that we can then use to infer what is in the mixture. So if you want specific sensors we need specific materials to bind to them. But if you want to know what's in the mixture we can do a lot with inos and machine learning. So the next questions will be to everyone. So what are the key challenges and opportunities you see in the current application of wearable technologies for environmental health and biomedical research. And how do you think they can be addressed or leveraged for better outcomes. Anyone. I can jump in and I would say that on our end one of the big challenges is the the battery life issue and so it's super exciting to hear the work going on at the assist center, because when we talk about adherence to device where one of the challenges is that people take devices off to charge them forget to put them back on. And so the longer we can have devices on the body, where people don't need to remove them. The better we will get with adherence the better data will get the better predictive algorithms we can build. So it's thrilling to hear what's going on at assist. I'll be very, very brief. By the way, not only with what assist is doing which is next level but also even, you know, wireless based charging, I mean, you know, through the wall charging even using very, very weak signals like Wi-Fi for things is super exciting on the power end. But I think one of the big challenges is interoperability and you know I think it's exciting that we have Apple here. And it was just spectacular talk from from from both truth and and I'm really look forward to seeing what Dr. Chung have to deliver in the future but having kind of platform agnostic technologies that could talk to one another, and then delivering things to whether it is someone about toxicities on the environmental side or to a clinician or to a patient or purchase a person at home. That is actionable is really what is next level to really avoid kind of a paralysis of analysis, because if we can do that with all this stuff flooding into us now. Those are the masters of the universe that are going to make the biggest difference in the future. Anyone have any final comment for the question. I would just add that in environmental sensing specifically there are perhaps additional challenges, just making sure that that environmental sample is getting to your device. And the surface is not getting contaminated over time, which is not just for wearables as for all environmental contact based sensing. That is a challenge that will need to be overcome, especially if you want to do long term monitoring. I'll just add one other aspect of timing the exposure to the relevant time window for the biological outcome I think is really important maybe not a stumbling block I think it's achievable but with all of these advances. It seems like there might be potential to really hone in on key time windows, and that's really exciting. What I'll add is I think there is a really interesting opportunity here highlighted by all the talks that we've heard today and it sounds like yesterday about the move from population health to personalized health right so taking all of that data that that we're coming from individuals where they are within their life and being able to use that data hopefully in a way that preserves a user's privacy that really helps us understand a little bit more about what is needed for a specific individual and helps individuals themselves understand their baseline and what isn't isn't normal for them within and outside of the environment. Yeah, thank you very much. Thank you everyone for a wonderful presentation and also Q&A sessions. So we will be concluding this session, and we will be now moving on to the next session panel discussion, moderated by Dr. Jennifer and Dr. Trey Thomas. Thank you very much. Hi everyone, is everyone able to hear me or any anyone great. Thank you very much. So my name is Jennifer Horney. I'm a professor of epidemiology and a core faculty member in the Disaster Research Center at the University of Delaware, and I want to welcome you all to session for on understanding how technology adoption implementation and science communication factor in advancing biomedical and environmental health research. I'm really excited to hear from our panelists. I don't know if we can top the last one, but we will try. So we will hear from five speakers as part of this. So I will briefly introduce them now so our first speaker will be Nita Farahani from Duke Law. And then we will hear from Dr. Chaturvedi, who is from the University of Southern California. And then we will hear from Dr. Bansali, who is a distinguished professor at Florida International. I would just like to remind everyone that there will be like in the past session on all panelists Q&A session so please submit your questions through the Slido during the presentations. And so I will without further ado, introduce our first speaker on this panel, Nita Farahani. She is a distinguished scholar on the ethical, legal and social implications of emerging technologies, as well as the Robinson Everett Distinguished Professor of Law and Philosophy at Duke Law School. And the founding director of Duke Science and Society, and faculty chair of the Duke Masters in Arts and Biomedical and Science Policy. She focuses on the implications of emerging neuroscience, genomics and artificial intelligence for law and society, and legal and ethical issues arising from the COVID-19 pandemic. She is also working on FDA law and policy, and the use of science and technology in criminal law. She has legal and scientific publications. She is also the author of the book, The Battle for Your Brain, Defending Your Right to Think Freely in the Age of Neurotechnology. In 2010 she was appointed by President Obama to the Presidential Commission for the Study of Bioethical Issues. She received an AB in Genetics, Cell and Developmental Biology from Dartmouth, an ALM in Biology from Harvard, and a JD and MA from Duke, as well as a PhD in Philosophy. And I actually heard her speak on NPR, so I'm very excited and would like to invite her to present her, begin her presentation. Thank you. Thank you Jennifer, and thanks for having me. I don't know if I can top the NPR episode I suspect that one was edited, but you know, I'll do my best. So, I'm going to focus specifically on the area that I know the most about which is looking at brain wearables, or wearables and the brain more generally and kind of the coming age of using sensors and everyday technology and what the implications are for privacy and what kinds of rules we need to put into place and what kinds of oversight we need to put into place and empowerment of individuals we need to put into place to really realize the benefit of this coming new area of sensors. So, right now there are and let me just make sure are you seeing the slideshow version. Okay, good. So, right now, there are some tremendous investments that are underway, both by neurotech companies but also by large tech or the kind of big five into embedding brain sensors into everyday technology. And up until now, most consumer wearables for neurotechnology have been quite limited, both in terms of the form factor because they're uncomfortable headbands that are unlikely to be worn by most people as they go about their everyday lives. And they've been quite limited in application to things like meditation or improving focus. So they've really been niche products with limited applications. But as companies from Metta to Snap, Microsoft and even Apple secure patents acquire companies and start to look at embedding brain sensors into everyday technologies and as the technology improves to enable EEG sensors to be put into two years with better powered algorithms that can filter out more and more noise to get better signal to noise ratio to improve what those signals can be interpreted and mean the possibilities become much more significant. Just today I saw another company who launched a product that is purportedly a neural interface device that I'll talk about in a moment, but the idea is, this is really, I believe the kind of final frontier of sensors which is to bring sensors to the brain and not just to the wrist for picking up heart rate or to the fingers for picking up sleep and other activity or EMG sensors that are picking up bioelectric potentials throughout the body. And the promise of doing so, I think is extraordinary. And that is the toll of neurological disease and suffering and society continues to rise, even as physical health and longevity continue to improve the implications for individuals to have so little insights into their own brain activity, objective given the other objective data that they have access to. I think in many ways holds us back and the possibilities of wide scale use of brain sensors and everyday lives creates the possibilities of creating large scale data sets of individuals engaged in everyday brain activity so that we just might one day be able to address the leading causes of neurological disease and suffering and even be able to get a better handle on our stress, have better interventions for things like ADHD, improve our focus in an increasingly distracted world, and even find the things that theoretically engage our brains the best, rather than the things that potentially diminish and make our lives actually net worse. One of the things that's interesting among the many is that these brain sensors are being looked at not just for increasing the transparency of the brain and making more accessible to individuals, but also potentially a serving as an interface to other technology, which is different in kind existing sensors and opens up the possibility of much more ubiquitous use of that technology. It also opens up the possibility that it will become deeply integrated into our everyday lives. Well it's still early days at really achieving neural interface as I mentioned I saw a company just today that announced a wristband that purportedly is neural interface that can be attached to an Apple Watch so that people can do things like interact with the rest of their technology, while having I thought initially looking at their advertisements that was the MG it looks like bioelectric potentials that could be used to try to do things like interface with other technology. The complications of the space of course is the fact that by most by now most people realize that free services or the services that are provided by big tech company comes at the expense of individuals, personal data and privacy. While Google originally sought to bring order to the web and provide high quality search results that now commands 92% of the search market that may change with the advent of chat box and LLMs. But tech companies business models have been resting upon the ability to sell their understanding of us to others. Google does this through real time bidding process which provides advertisers with opportunities to uniquely target advertisements in real estate. Meta does much the same thing harvesting data on billions of users and creating psychological profiles of them that advertisers can use to micro target their pitches. Shoshana Zuboff coined this term surveillance capitalism to describe this now ubiquitous phenomenon characterizes data about behaviors of bodies minds and things and surveillance assets that can be used for purposes of knowing controlling and modifying behavior to produce new varieties of commodification, monetization and control. And I would argue that our brains are truly our last fortress of privacy that they are the kind of one space in which there is some at least unexpressed emotions unexpressed thoughts unexpressed words and images that can't be easily detected through other kinds of inferences. And while again it's early days still and most people would argue that the kinds of data that you can gather from so many other aspects of our lives are going to be more important than the data you gain from the brain. That will change over time and the ways in which it's already being misused in employment contacts by authoritarian regimes who are requiring or forcing the use of brain sensors who are using it to threaten to coerce requiring students to wear brain sensors in the classroom to track their attention and focus. This can have a profoundly chilling effect in ways that can undermine rather than empower individuals. Nevertheless, I strongly believe that we have to get to a place where there is the possibility of sharing data. And the only way we do so is by creating a system by which people can be confident that the use of these sensors are empowering to them rather than disempowering, which means we need to put into place. I believe a right to cognitive liberty and that right to cognitive liberty I think is both a human right, which protects people's right to self determination to access and use such wearables, but a right to mental privacy and freedom of thought, which means that the default to brain wearable shouldn't be the same as social media platforms but treated as more sensitive data that multifunctional devices should have greater user control. If you're using your earbuds to listen to music there should be times at which you can turn off the disabled features of things like brain activity tracking. And we need to determine how to put safeguards into place for individual and group harms that can arise from the sharing of this data to enable people to confidently use brain sensors, and to safeguard themselves against the potentially deeply intrusive privacy risks that can arise as a result. Thanks and I look forward to the conversation. Thank you so much, Dr for honey, and hopefully I'm pronouncing names correctly. Our next speaker is ratika chatter Betty. Dr chatter Betty is a biomedical engineer at the USC Schaefer Center for health policy and economics. She has a diverse background in engineering science and technology policy asset valuation strategic consulting and translational biomedical research. The research involves improving fairness and equity and precision digital help. This really leads the American life in real time study, which involves Fitbit collection from a large representative sample of participants established using fair standards, findable accessible interoperable and reusable. This research serves as a benchmark cohort and data set for the digital health community to improve transparency and validity in digital health model development. So we'd like to invite Dr chatter Betty to begin her presentation. Thanks so much for the introduction and for having me here. I'm thrilled to introduce our American life in real time study to this group. As it was mentioned, earlier was designed as a benchmark data set for equity and digital health and it's an interdisciplinary partnership spanning health policy economics engineering and computer science across USC and our partners at evadation health. We got a lot at this meeting about the benefits of wearables and digital health and how we can create personalized health interventions using large scale data and advanced analytics so I won't go into too much detail about that here. But what I wanted to focus on today is that one of the benefits that's often cited in this work is that machine learning and big data have the potential to reduce both explicit and implicit biases and life science research and medical decision making. What I wanted to emphasize is that there are reasons to be very skeptical of this claim. So this is what the digital health framework should look like in theory, but in reality it looks a little different. And it turns out that there are field wide methodological biases that systematically under represent minoritized populations. First, most studies rely on convenient samples of easy to reach populations and bring your own device designs. This creates selection biases in the data from which these algorithms are trained. Second, advanced analytic approaches often rely on the use of complete cases where data with a lot of missingness are eliminated from the analysis. However, we know that device use and competency is strongly associated with socioeconomic status. So this creates a one to punch. You have a non representative population that becomes even less representative during the analysis phase. So how bad is this problem. If you look at the understanding America study, or UAS, which is a probability sample of US of US adults, only about 22% of the sample are self motivated device owners, which is in line with estimates from others such as Pew. But what I want to emphasize is that the socio demographic distribution of these individuals differs significantly from the US population. Here you have the general population in black, the US UAS sample in white and self motivated device owners in green. And you can see that owners are significantly more female, more white, more educated and with higher incomes. Consequently, excluding non owners from digital health studies results in demographic and socioeconomic disparities across almost all factors. And these demographic distributions even just differ based on the type of wearable here Apple watch owners skew even younger, more educated, more likely to be working and with higher incomes. Now this is important because most studies will choose a device to be able to compare across people. So what is the gap in the field. It really is that benchmark data sets are urgently needed for the digital health community to test train and validate AI and ML models in an equitable and rigorous manner. And there are a couple important concepts here. So first benchmark data sets are used to establish performance criteria for software tools and afford more transparency in a field that inherently deals with algorithmic boxes. Second, fair standards which stand for a findable accessible interoperable and reusable is an emerging standard in computer science to improve the democratization of big data analytics. By combining these principles we came up with a list of design criteria for benchmark for a benchmark PG HD data set. First, we wanted to ensure socio demographic representation and came up with four criteria, national probability sampling with a focus on inclusive representation of marginalized populations, hardware provisioning, encompassing both internet mobile devices and wearables. And then second, we wanted to ensure rigorous labeling, which allows for machines to learn patterns in the data. So long term longitudinal data collection to capture temporal patterns, and well validated measures through surveys, which allows for low cost repeated collection, when compared to clinical or biomarker endpoints. There are quite existing studies in this space. There are only a handful of studies that meet the fair criteria, namely and Haynes, the UK Biobank, all of us, and the framing ham part study. And they fulfilled the benchmarking criteria to varying degrees, but none of them really hits all of the criteria. And this is where the American life in real time comes in, in partnership with the understanding America study, we aim to achieve both socio demographic representation and high quality labeling. So how did we design this. So first we enrolled a random sample of about 1000 adults from the understanding America study which again uses address based probability sampling to recruit its cohort. Participants who did not have internet are provided with a tablet with forgy connection. We provide everybody with a Fitbit device as an incentive, regardless of whether or not they already have a device. Participants contribute periodic surveys through their involvement in both UAS and all year. And we also match contextual data both temporally and geospatially. The infrastructure is flexible enough to accommodate future data streams from literally anything with an API, including electronic health records genomics biomarkers apps and even medical devices. And we follow these participants for as long as they're willing to contribute the data, but at least for one year. Let's look at the cohort itself. So our main question here was, through our methods, were we actually able to achieve socio demographic representation. And the short answer is yes, here you have population benchmarks in black and the unweighted alir sample in green. And we find that our unweighted data set is largely representative across as characteristics. Now when you compare this again to self motivated device owners in orange here you can see how stark the differences can be. We only balanced our invitations on socio economic factors such as sex, race, age and education, but in recruiting a representative population based on SCS alone, we were pleased to find that we also achieved a largely representative sample in terms of health characteristics as well. And again, if you compare this to self motivated device owners, you see that they're much more likely to be significantly healthier than the general population. So, next let's talk about what's actually in the data set. And just to give you a sense of the multi level late nature of the data. The light green boxes here represented by annual data data that we have for these participants, including the full health and retirement survey instrument which goes back to 2014 for some participants. The light blue boxes are the monthly longitudinal surveys we added specifically for the study, primarily in the social structural and environmental determinants of health and health measures that are likely to change month to month like physical and mental health. The dark green boxes are derived measures from the minute level sensor data from the Fitbit API. And finally the dark blue boxes are contextual data from public data sets such as pollution, crime, whether that we match both geospatially and temporally to participant timelines. So finally, let's talk about why this matters. And I'll give you a very quick example with COVID-19. We found that there are significant changes in Fitbit biometrics from individual specific baselines during COVID infection. And this figure shows the change in heart rate, but we say changes in activity, sleep, walking speed and several other important Fitbit measures. If you throw all these data into a machine learning model, we find that we can predict COVID onset and disease with pretty high accuracy. But let's look at how sociodemographics influence model performance and bias. On the left, you have a model trained on a representative sample from earlier. On the right, you have a model trained on just self-motivated device owners from the UAS cohort. On face value, the performance for both of these models is more or less the same with an area under the curve of 0.85 on the left and 0.83 on the right. However, when you test these models on out of sample individuals, you see that the BYOD model does not perform as well as the Aleer model in underrepresented groups. Here we're showing performance in males as our BYOD cohort heavily skews female and you can see that the performance drops substantially on the right. And you can see similar trends in non-white groups and lower educated groups. So collectively, this demonstrates the need for representative sampling for equitable model performance and AI machine learning. And so this brings me to the end of my thought talk. I'd like to thank our growing team of supporters across multiple universities and our sponsors at the NIH. And thank you so much for having me here. Thank you. I think we definitely could have had our whole session talking about the interesting findings and those research, those curves that you showed at the end. So I hope that everyone is thinking and submitting their questions. So our next speaker is Dr. Bonn-Sully from Lucent Technologies. The distinguished university professor is a Lucent Technologies Distinguished University professor at the Florida International University. This is basically the director of the U.S. National Science Foundation Division of Electrical Communications and Cyber Systems for two years. Dr. Bonn-Sully's expertise is in the field of biosensors, microfluidics, nanostructured catalysts and microsystems. Dr. Bonn-Sully also has a background in developing microfluidic tools for DNA damage detection, 3D multicellular sphere monitoring, real-time biomarker monitoring, automated cell help, cardiovascular diagnostic sensors, and many other diagnostic devices. He is recognized for mentoring through multiple awards, including the Alfred P. Sloan Foundation Mentor of the Year and is an active member of the Institute of Electrical and Electronics Engineers and a fellow of the American Association for the Advancement of Science and the National Academy of Inventors. We would like to invite Dr. Bonn-Sully to begin his presentation. All right, so hoping you can hear me. Yes. Excellent. Okay, so let me go back. So what I was talking about, and it's fantastic talks, right? So I was trying to switch gears and look at both technology and the workforce that drives our work, right? So like I said, we started as my career, doing MAMS devices and making this really cool stuff and lab on a chip things, and then eventually transitioned into wearables. So what I'm showing you here is one of those wearable wound sensors, what I have to put wounds that we did in collaboration with Vina as part of the assist center you heard from me earlier. And this was basically us monitoring a couple of headlights there, right? And then we did the same thing. Cortisol is an example that was talked about yesterday. We did the basic test but then developed the basic technology for point of care use, point of use. And it was used both for looking at farm workers and their toxicity exposure with one colleague and the patient and pregnant women with another colleague, right? So the same biomarker depending on the circumstance coordinated to clinical. So, but if you take a step back, and we look at what we really are trying to do at the intersection of disciplines. And this is like more electrical engineering forum and I have to have a Mickey mouse. I'm from Florida, right? So, but if you take a sense or stick it to the body or anybody working on it needs to have a good handle on multiple domains. And I'm just missing some out and I'm sure I've missed many, right? And at least an awareness of many others. So if you're trying to do this, how do you accomplish this? Especially in undergraduate and graduate curriculum. So, so the one thing that I wanted to share with you is something that when I was at USF we started was this whole idea of interdisciplinary training when we're still very nascent NSF I just started I goods, right? So, between USF and FIU, the institution spent about $2 million total of the institutional cash. And we started our programs at intersection of where there was discipline, the first one at USF was doing marine science and engineering, right? And the critical part on trying to get the idea of going back and forth was actually co-locating students and making them just sit in the same shared space. And that seemed to work, right? So what are the last 20 years that this just program that I'm involved with? We went from that initial $2 million investment in two institutions to over 200 minority students getting funded so far. And you know collateral to that is things like the ERCs at which FIU is part of three. I'm part of the partner with us at VINA. And that we don't even count those research grants, right? So institutionally, I think there's a huge ROI as investments are made. But we take a step back and look at a curriculum structure because I don't know how many of us really carefully follow what's happening in academia and where the trends are. But 10, 15 years down the road, our supply line is not going to look anything close to what we have right now. Our undergraduate curriculum is 120 courses, 60 classes of gen ed, 60 credits of per division, very traditional. So how do you get the depth that they need and that the breadth we need for these kind of works? So something that we've played around with that FIU when I was a department chair was we just change a curriculum. We went from six electives undergrad to 32 electives credits, you know, track based curriculum, opening up a graduate curriculum, and then have specialized degrees, you know, which are kind of going in depth in the same field. But increasingly looking at immersive learning experiences, I mean, we have a very thriving undergraduate research experience program during the year. And with the very broad participation of students, right? And I think that's one way that this will work. The other challenge that I think I see for a lot of work we do is I just just googling up what are the, you know, last 10, 15 years, what are the coolest thing that happened. 2021 is a start because that's just my world. I mean, for those of you don't know the gallery test. This is a blood test that was, I think, double by company that speeds you for 50 cancers with pretty high accuracy. But if you look at it, a lot of the stuff that we do underlines in terms of technology underlines the products that are being that we're looking at. But it doesn't translate into excitement, bringing back students to basic sciences. So it's it's the BS, a BSF code, right? How do you make sure students understand that we, we are behind every product that's being made out there, right? And that becomes that I think is a challenge for us and continue to be a challenge and how we get the workforce and the pipeline to do the kind of work that we want done. This is where we are. I mean, if you look at the hype curve, everything we are talking about doesn't make it to the hype curve. So, given conversations and interests are driven by social media, how do we, you know, get popular. So I wish I had answers. I'm just kind of leaving open questions as to how do we attract the best talent into our areas. And that remains a challenge that we increasingly wrestle with. So, you know, coming to the end of my time, so I'll stop here thanking my group and my sponsors, both current and former with that I'll pass the mic back. Thank you very much, Dr. Boncelier. I now have the honor of turning the rest of the session over to my co-moderator, Trey Thomas. Good afternoon, everyone. My name is Trey Thomas. I'm with the US Consumer Product Safety Commission, and it's my pleasure to serve as the moderator for panel two for the panel two presentations. Unfortunately, we had one of our panelists had an emergency, but we will continue on to hear from Dr. Tiffany Powell Wiley and Deborah Prince in these last two presentations. So I will go ahead and begin to introduce Dr. Tiffany Powell Wiley. She is an Earl Statman investigator and chief of the social determinants of the obesity and cardiovascular risk laboratory at the National Institutes of Health. She has a joint appointment in the cardiovascular branch of the Division of Intramural Research at the National Heart, Lung and Blood Institute, and the Intramural Research Program at the National Institute on Minority Health and Health Disparities. Dr. Powell Wiley's interdisciplinary team uses community-engaged research epidemiologic methods and translational approaches to better understand social factors that promote obesity and limit cardiovascular health. In 2023, Dr. Powell Wiley was chosen for the American Society for Clinical Investigation, and Dr. Powell Wiley graduated with a degree in engineering from the University of Michigan and has an MD from Duke University School of Medicine. So with that, we'd like to invite Dr. Powell Wiley to begin your presentation. Thank you so much, Dr. Thomas, and thank you to the organizers for the opportunity to be here today. Today I'll be talking about using digital health technology and community engagement and how we're using digital health technology to promote cardiovascular health equity. These are my disclosures. In terms of the promise that we know relates to the use of digital health technology, particularly in addressing cardiovascular disease disparities and cardiovascular disease prevention, we know that we can think of the use of technology really across the spectrum of prevention. And that can be from primordial to secondary prevention where we can use things such as telemonitoring, text messaging, mobile apps, all of which can play a role in knowing tailoring information and tools for users, particularly those patients who have limited access to health care. And we also need to think about wearables and how those can be used along with many of the sensors that have been mentioned as a part of this meeting, and not only measuring environmental exposures, but also psychosocial and stress related factors that we'll experience by those populations that are most impacted by health disparities. And so in thinking about utilizing those tools, we also need to consider potential barriers that exist, whether it be cultural barriers related to trust in the health care system and providers that promote the use of those tools. Concerns about privacy and safety of health information, but also concerns about potential discrimination based on data that is collected and all of that limits or impacts the willingness of individuals and patients to use digital health technology. And so we've put forward that community engagement is really key in thinking about how do we build the partnerships with community populations, again, most impacted by health disparities to really gain feedback through the process of developing interventions. And so we, I put forward in a recent perspective and circulation of how we really need to center patient voices in the development of interventions, more broadly in cardiovascular research, but we can think of this the paradigm that is in the process that are involved in community engagement and really developing interventions that not only help in improving cardiovascular health but also in addressing social determinants that impact cardiovascular health. And so if we're going to develop interventions that utilize digital health technology and engage communities, we need to make sure that these interventions are accessible to all and really think about providing wearables and devices as a part of interventions but also think about making sure that these studies are built on the trust within the community that they are, they built in a sustainable manner but also that they utilize and recognize that all communities have assets that can be leveraged to improve cardiovascular health. And again, many of the digital health tools that have been mentioned here but can be utilized in promoting and better understanding cardiovascular health can also be used to that to understand the role of social determinants, particularly like a social stressors that impact cardiovascular risk. And so in the work that we've done in Washington DC and building partnerships with my group at NIH but also groups at local universities including Howard University, as well as community based organizations that do work around improving cardiovascular health. We've also brought to the table, community members as a part of a community advisory board, called the Washington DC cardiovascular health and obesity collaborative, or DC to work in developing and providing feedback on each step of the process for the development of intervention and the culminating intervention that we are currently implementing from working with this advisory board is called to step it up physical activity intervention. And so this community advisory board of building partners from all of these institutions locally has met quarterly for almost the past 10 years to provide feedback on each step of the process and developing the work that we do and so they not only provided feedback on the design of the mobile apps that we're using as a part of our intervention, but they helped to help us understand how these apps could even be utilized for measuring cardiovascular health in the community. And we also have worked to really work within the community to understand how we can disseminate the findings from the that what we are learning gets directly passed back to the community so that we can further tailor on future interventions, or that we can work with community and academic partners to develop more detailed focus interventions based on what we're longing about for cardiovascular health. And many of the tools that we've used have been really have been designed using a user centered design process when we worked with community members to do user experience testing for each of our, the digital tools that we're using in the, as a part of the and so just to take a step back and to show kind of where things started. We started by really recognizing the disparities in cardiovascular health and really looking at where risk for cardiovascular disease was most prevalent in the city and we targeted those areas in the Northeast and Northeast DC, where we knew obesity was most prevalent but we also knew that cardiovascular disease was highest in these communities. These were areas most impacted by disinvestment and segregation. And so we focused on doing data collection at community sites to not only understand what cardiovascular health looked like in the communities, but also to look at utilization of tools for measuring and monitoring cardiovascular health using wearables as well as mobile apps within the population. We identified physical activity as a particular target for intervention, particularly among African American women in the population. And we were one of the first studies to show how wearable physical activity monitors can be used in the population we were working with who are at high risk for cardiovascular disease. And so it allowed us to show the feasibility of using wearable devices within the community for future intervention and really set the stage for the design of the step it up intervention, which is a sequential multiple assignment randomized trial or smart study that uses mobile health technology to promote physical activity for African American women living in our target areas of Washington DC and surrounding areas of Maryland. And so this is a intervention that's built on a multi level framework. It's designed to use digital technology to test whether messaging that's been developed can change for individuals perceptions about resources for physical activity within their environment and whether that improves physical activity over and above standard messaging. And we developed this intervention over a phase approach where we have a first phase where we designed the metrics for the study. The second phase was extensive pilot testing. And right now we're in our efficacy trial where we're implementing the randomized trial and this was built on the orbit model for designing behavioral interventions. So just to give you a sense of what the how the messaging works for the intervention. Again, it's built on a multi level framework and we adapted the social ecological model to develop messaging for the intervention and so we have messaging that works at the individual level that is more focused on getting individuals to build, gain social support and set goals for physical activity and work to maintain those goals as a part of the intervention. This was developed in collaboration with investigators at GW. And then the second type of messaging is more built on the to work at the neighborhood level is called location based messaging. And this is designed to work by using geofences to provide messaging about physical activity resources. When a participant is near within a certain distance of those resources. And so we worked with community members to identify community assets for physical activity that we might not necessarily see on maps or websites. And so as we design this test these tools and this messaging as a part of the mobile app for the intervention. We tested each version of the app using both qualitative and quantitative methods to test the use of the app as well as barriers and facilitators to app usage. And we incorporated suggestions for improving the app into the next design. And then we tested both the standard messaging as well as the location based messaging through pilot testing and now we're, as I said implementing the full six month physical activity intervention. And so this is just one example of how community engage research can really be used to develop these digital health interventions that are tailored to those populations that are most impacted by health disparities. And really we can think of the important goal of digital health technology and addressing social determinants that influence not only cardiovascular disparities but really chronic disease disparities and ultimately by incorporating this type of data into these into community engaged approaches, we can not only develop more equitable interventions but also as was mentioned previously more equitable algorithms for the work that we do. And so I'd like to acknowledge my partners and collaborators and those who've worked with me on this work, especially our community advisory board members, and I'm happy to take questions at the end of the session thank you so much for the opportunity. Thank you, Dr. Paul Wiley that was an excellent presentation and really glad to see the application of the wearable technology. So thank you for that. So now we'll turn to our final presentation, and that will be Deborah Prince, and I will go ahead and read her bio. Deborah Prince started her career in standards with you underwriters laboratory or you will in 1995. She worked in standards operations, then in global standards where she was involved in implementing the STP standards development process. Currently she is the director of standards programs. These programs include augmented reality, virtual reality, medical devices, software safety, cybersecurity, and the ULSEs autonomous suite of standards, which include automotive heavy trucks unmanned aerial systems, LiDAR and robotics. Deborah holds a BS in mechanical engineering from the Missouri University of Science and Technology. And with that we'd like to invite Ms. Prince to begin your presentation. Thank you, Dr. Thomas I appreciate the opportunity today. So, listening to the discussions that have come before me, the conversations have been around wearables, and a variety of issues with wearables. I'm going to come at it from a different lens and talk about wearables, and how do you know that they're safe. And this is particularly in the fact of virtual reality, augmented reality, which makes reality technical equipment. So, UL standards and engagement has a standard that has recently been published called UL8400 that addresses these issues and risk and providing a platform that you know the type of risk have been addressed. Right. And, sorry. Okay, yeah, we can see your slides. Great. Welcome. When it's been a second letting everyone know who you all standards and engagement are. So we are a company that is we like to say we are safety science and action. We are a global nonprofit. We developed our first standard over 100 years ago. It was 10 cloud doors, it's still being used today and it was published in 1903. We have over 700 standards using an open transparent process, we have over 1700 standards in our portfolio. We have staff of over 100 dedicated staff, we're growing rapidly that number is probably closer to the upper hand of the hundreds right now, we are located in eight different countries. We have participants in our balanced committees from 600 or 60 plus countries. Of those consensus body members, there are 4,000 or more. An additional bank of participants that are, we call stakeholders who like to follow along but they may not be on a committee to vote, and we have over 50,000 of those, and we have over 400 individual members are committees in existence. So let's move on to, why did we develop a standard and how did we develop the standard to address ARBR MR. We were approached by CPSC. And they had had a lot of incidents reports on injuries. And they also had done a summer study. And a lot of these injuries were situations of falls, not knowing the surroundings, and then a variety of other issues. In February 2020, we started the work on what will become UL 8400 we formed a committee, we move forward. In March 31, we had our first in 2020 we had our first in person technical committee meeting to be scheduled. However, as everyone knows, mid March COVID shut down the world. So instead, we had a ton of virtual meetings, where we went over a first draft. We had a lot of test test groups we went over a second draft. We had some more meetings some more tasks groups. And finally, in March, the end of March, you will 8400 was published. It's a national standard in the US and Canada that can be used anywhere in the world. What's in 8400. Well, first of all, let's talk about that it is not something that can be used standalone, because the risk that are identified there are not all the risk in a product. So it needs to be used in conjunction with a standard such as can CSA. 62368-1 or UL 62368-1. A lot of numbers there but basically what that standard, those two standards are, they're harmonized are consumer electronics standards so they're going to give you the electrical risk, the any kinds of electronic risk, and make sure you have a baseline of safety of fire, shock and safety handled. Now there's a variety of different ways equipment is used in different products in the marketplace so one way to address all the different types of products is that you'll have to use a risk assessment in the standard. And like I said, it was developed using an open process using the committee that we call our technical committee. So the technical committee had representatives from a variety of countries involved. So, the standard looks decided how to categorize the different types of products that are out there. We decided to look at non see through. So you've got a hazard in that hardware non see through means there's total optical occlusion. So again, you can't see through it video see through so it's opaque but it has some video pass through the cameras and then optical see through so most of the users vision is still maintained. So what's covered by ul 8400. So we have see through visual functions. So that's going to be covered and needs to be addressed during a risk assessment, most of that those functions, what kind of ambient light levels are intended to be used with the products, or can it be used outside can it be used inside can it be in a dark room or not. Also they need to address the kind of use cases if there's poor visibility, such as an industrial environment. The headset being used and maybe there's heavy equipment nearby machinery that could cause a risk to the user. And then any kind of risk that come into effect from the lens filters on visibility. So the standard also addresses flicker. So flicker is, we're worried about any kind of biological effects induced by strobing. And the standard requires that that's minimized to the extent possible you don't want something that would trigger epilepsy for say, or something like that. Compatibility is important. And really the concern there is prolonged skin contact. So you need to make sure any of the materials that come in contact with the skin again for a prolonged period of time are made of something that will not cause a rash cause burns cause any harm to the skin. In addition, we need to look at the standard also looks at and requires that the recommended cleaning practices are addressed. And recommended by the manufacturers cleaning process knowing that you can't control how somebody cleans a product but how they're recommended to be cleaned is important, and that those products do not affect the skin that is being used and cleaning. You also need to look at exposure of the eyes to thermal energy. So that's heat, you can't have a headset, providing too much heat that can damage the retina, the lasers so you look at lasers and other stresses there. Biometric stress so what you're really looking for there is any problems with the upper cervical spine. So compression. When I get to that point, I guess a good point here to bring up is that the standard is you intended for anyone over the age of 12. There's really not science there yet for us to have good numbers on what you would do for a child. We're hoping in first further revisions there'll be more data and we can address younger ages with the standard. So, and also within the biomedical stress, biometric stress, sorry, no biomedical, you look at some mechanical rough robustness so there's tests to break the glass and make sure there's no. It doesn't mean that the unit won't break is dropped it's simulating misuse perceivable misuse and it doesn't mean the unit will break. And it is making sure that it doesn't break to the fact that there could be harm to the user like cutting their eye or cutting their face. In addition, we want to make sure when you're doing these drop test and verifying robustness that there's no harm to the battery pack these are using lithium ion batteries. Those can get very hot you can damage the unit. Any damage to the unit when charging could potentially cause problems with fires and different things like that so we need to make sure that these units are very robust. Right, we have the enhanced spatial perception so you're looking at whether something is mobile or static and how you use it do you know your boundaries different things like that. There are standard addresses safety and warning instructions as well as functional safety so does the unit function is intended and is it safe. Deb I hate to interrupt we have about a time until we have about a minute if you could wrap up quickly. Thank you. Absolutely. This is pretty much my last bigger slide so morning and labeling all of this is done through a risk assessment. You need to look at basic health and safety information preparation of a safe environment preparation the device safe use of the product and accessories maintenance and storage. Your vulnerable age groups. So you're elderly. Again, we're not talking about 12 and under so they're not and then a first use tutorial. And last but not least what does the future hold. We're already under continuous maintenance which means we're getting back together we're going to advance the standard to address more things. And other risks that haven't been put in there yet. I'm also going to look at do we need to expand the scope do we need to look at other technologies as this continues to expand things like that. So, I want to leave you with if anyone is interested in participating on that please reach out and love to get you engaged. All right, with that, I will go ahead and turn it back over to Dr Thomas and thank everyone for the time. Thank you so much, Miss Prince. That was an excellent presentation I think shows the, the other side of the technology the importance of standards and developing good safety practices so thank you very much for that. Thank you, everyone that you can provide questions into the Slido, and it will go ahead then I think that's horny and I will trade off with the questions. Jennifer would you like to start with the first one. I was going to say you've got the mic if you want to go first or. Okay, let's see. Okay. This is to Dr Tiffany Powell widely. How effective is the tailored message intervention for your community folks to increase exercise. Do you have specific outcomes data. Do you have economics analysis data. Do people stop responding to texts after a while. We are still analyzing data so we're not near the completion of the study so I can't give all of that information we did pilot testing that show that the tailored messaging location based messaging increase physical activity. Over the over the standard messaging on it would approach statistical significance but didn't reach statistical significance but we're that the goal of this, the full study is to test that that is our primary outcome. And of course we would want to see the cost effectiveness of the intervention as well, but we are very please stay tuned that's my side. Thank you. Thank you for that answer. Jennifer, do you want to ask the next question. Um, so it looks like we had a few questions about the concept of cognitive liberty and maybe some discussion about how that may relate to children and there was some some conversation in the in the chat but I wonder if we could maybe have a little bit more discussion about how that relates to empowering parents and diagnosis and other tracking for childhood biomarkers and neurological. I don't know I weigh in a little bit on that Jennifer so um, so part of how I've been talking about cognitive liberty in this context is to really recognize that it's like Liberty as we understand it and the digital age and the age of centers and wearable technology isn't just about a right from interference with privacy or a right from, you know, security against the kinds of discrimination or misuse but it's really about a right to self determination which includes a right to informational self access, a right to be able to use that information to improve ones and you know the context of brain brain health and wellness and, you know, capability of decision making and so if we're looking at children, you know, this matters and thinking about their right to cognitive liberty their right to be able to develop cognitive freedoms capabilities, which includes mental agility, relational intelligence which includes emotional intelligence and their capacities and competencies. And so if there, you know, is in the pediatric population the struggle over, you know, the, the use of wearables for example and the potential, you know, implications from a privacy perspective it's also really important to think about the cultivation of those capabilities in children, those capabilities especially in children are critical to their ongoing development to their ability to further, you know, engage over their lifetime in self determining choices. And so, you know, in this context thinking about that I think it's really about how do you ensure that you know if there are necessary set of stimuli for example, if they need to be able to hear better to be able to see better to be able to process information or have different kinds of stimulus to enable that cognitive development that those will be crucial inputs to their development to cultivation of cognitive liberty. And so, you know, this is what I've been working on a lot is thinking about not just what does that mean at the human rights level but what does that mean for cultivation, you know, from well beyond the regulatory space but investment in research that enables that investment in policies and ethical guidance that would really focus on doing so whether it's in children or in adults. Great, so I'll ask one more trade you don't mind so this was related to some of the information about covert surveys leading to different predictions for different groups of people and comparing the two models. The patient replies differ between represented and underrepresented groups. And do you think that the two groups had different coven symptoms. Yeah, so that's actually a really great question so there's a couple different ways that we can look at the responses so obviously the Fitbit data that underlie the coven. is objective but we did ask a self reported question of whether or not somebody was diagnosed with coven and what the date was. So, there's a couple interesting things that we found so in our representative population of about 1000 people over a timeframe of August of 2021 to April of 2022. There's about 300 coven positive cases in 1000 people in that timeframe, which is actually a very a pretty high sample size for a lot of these digital health studies so if you look at a lot of the other literature that has been published in the space. You see sample sizes of 200 to 300 out of about 20,000 people or 30,000 people recruited to get a similar sample size. Right so that shows first of all how a representative sample size can give you a higher disease prevalence because you're fundamentally looking at a more representative population right rather than a population that is more highly educated and therefore potentially staying at home and being able to work from home. So that's point number one. The second point is that we had a much higher representation of obviously of underserved individuals in our sample we had we had a higher proportion of people who are infected or African American a higher proportion of people who are infected who worked in homes that did not allow them to work from home, whereas in a lot of these b y o d samples, not only is that data not reported, given the intrinsic composition of the sample we would not be able to get at that. So, we did see differences and who was getting coven in our sample number one, and number two, there are differences in the signal so for example, we saw that certain groups for example women had a higher had had bigger differences and for example sleep patterns and men during their coven infection so we're trying we're in the process of teasing those mechanisms out, but there are differences in phenotype by socio demographic status as well. Great, do you lose your. I'm back. There you are. Okay. Yeah, I'm having some connectivity issues but we'll march on. Would you like for me to ask a couple of questions we have a couple of good ones here. Well one, I think it's for could be answered by several panelists. And it says client data from wearable sensors should be recorded continuously, and over the lifetime of all individuals involved in the given corporate data so this is actually I think more of a given given core work so it's actually starting with data collection data should be stored for potential future analysis. How do you envision meeting the challenges of a multi decade data collection and storage and privacy issues, and this is really come up I know, including but not limited to personal health information GPS location relationships among cohort participants. Very interesting questions so what anyone like to take that on. I'm probably happy to jump in on that which is, you know, I think obviously there are huge questions from a storage perspective which I won't touch on but thinking about it from a privacy perspective. It raises significant challenges over the long haul, not just, you know I think it's both important for the individual as well as for, you know, the value that will come of the data and the kinds of insights that we can gain as a result. It increases the risk of longer term harms both in terms of kind of additional things and insights that you can gain from it or you know if you take it from a brain center perspective, looking at things like the rate of cognitive decline over time who has access to that how does that kind of policies does that inform is it used by employers, is it used as part of wellness programs and collected in the context of wellness programs is it's, you know, used to make discriminatory and other choices. To me, given the likelihood of re identification of most of this data, particularly as the, you know data becomes richer and more associated with other data sets over time. So I think that we think about securing against misuse rather than securing its access because I think access restrictions to the data, or kind of a short term solution to what will become a long term longitudinal data set that will open up kinds of different kinds of risk over time. And so my hope is that that's what we'll really focus on is trying to identify proactively and regularly what the potential risks of discrimination and misuse of those data are and by which actors and then to put into place rights and remedies to protect against that. Great anyone else for that question. I'll follow up that that's also very interesting. How do you overcome trust issues with wearable devices in vulnerable populations who are concerned about data being sold and are used against them in some way, for example, higher insurance costs, etc. All right, for whatever reason I lost the first part of that conversation it looks like critique I wanted to take that on did you want to take that on. I'm happy to start actually so in our experience we found that the major barrier to wearable ownership and vulnerable and underserved populations is not necessarily privacy and data security issues but it's primarily access. They're they don't know about what wearable devices exist they don't know the benefits of wearable devices and so they don't engage. And we find that if we provide education and we provide the wearable device free of cost that that reduces most of the barriers and we don't find that people from vulnerable backgrounds participate at any rate less than than the general population it's it's pretty focused at that point. However, beyond that, for the people who do express privacy and data concerns it's really about establishing relationships. In our experience we find that in our experience with establishing this panel over a decade a decade and having an ongoing participant narrative, providing feedback to participants about the usefulness of their data, engaging them in why these research questions are important, especially for people from underserved backgrounds really goes a long way and instilling trust. But that's just our two cents from our experience. Again, with our experience, we've done focus groups and as far as part of our pilot testing just to understand how mobile apps and wearables are used or what interests exist and digital health interventions in the communities we've worked in. And I would echo that the interest is certainly there. And it really is about it's not even necessarily about access to divide to the phone of course it's more access to the wearables and and learning but it's and it's not. It's really learning how they can be used and coupled with their phones and things like that, but the interest in using mobile apps and particularly using mobile apps for health related issues is there and it's happening we're doing that it's just how do we incorporate more digital tools in there, but the populations we work with as a part of our intervention are really excited about being a part of these types of studies and doing and being and having these types of studies focus on African American women for instance. So, definitely get a lot of that people. The answer is, unfortunately, this has been an incredible session and I, we really appreciate all of your insights. This is fascinating work. But unfortunately we'll have to wrap up Jennifer do you have any final comments before we turn it over. We're ready to introduce the next session which is session five on the future direction of wearables, moderated by our very good leader here of the committee workshop, and we'll turn it over to her. Thank you so much, Jennifer for all your hard work and for your generous remarks. I'm very excited to open up session five just as a brief reminder and introduction my name is Rima hover. I'm an associate professor in environmental health and at the spatial sciences Institute at the University of Southern California. I'd like to welcome you to our fifth and final session of this amazing workshop on identifying the research gaps limitations and future directions for wearables in environmental health and biomedical research. I'm also very grateful to our two invited guests, Dr. Joseph Wang of UC San Diego and Dr. Ed Ramos of scripts research and care evolution to have joined us in this final discussion session and future lifting session. I'd like to briefly introduce Dr. Wang before we give him the floor for a brief lightning talk, and then do the same with Dr. Ramos and then we'll have a great discussion at the end. So please be prepared with your questions when we get there. So briefly, Dr. Joseph Wang is a distinguished professor as I see in that professor at the Department of nano engineering at University of California San Diego. After holding a regents professor and last chair positions and MSU, he moved to ASU where he served as the director of the Center for bio electronics and biosensors at the bio design Institute. He joined the UC San Diego nano engineering department in 2008. He is a member of the US National Academy of inventors and of the Turkish National Academy of Sciences. Dr. Wang holds honorary professor titles from eight different universities and is the recipient of two national American Chemical Society awards for electro chemistry and instrumentation and the Chemical Society Sensors Achievement Award, the Charles and Riley Siak Electroanalytical Award and the Ralph Adams Hittken Award in bio analytical chemistry. He served as the founding chief editor of the Wiley Journal Electroanalysis and on the editorial board of 20 other journals. Dr. Wang is also a fellow of the RSC, PCS and AI MBE. His research interests include bio electronics and biosensors, wearable sensor systems, nano motors and micro robots and flexible materials. He has authored over 1200 research papers, 12 books, 55 patents and 35 chapters and he was ranked as the most cited electrochemist in the world in 1995. He is the most cited researcher in engineering during 1995 till 2005. And with that I'd like to welcome Dr. Wang and ask him to please take the floor for his brief lightning talk. Thank you. Thank you Rima and Natalie and thank you all for staying so late. So I'm from UC San Diego and the Pacific Ocean in La Hoya and we'll talk about the wearable sensor for environmental security. We were working on this for the last decade on the chemical sensor so we know that the topic started with a physical sensor for mobility vital sun, what we have in the market. What we would like to add is to introduce molecular information, go beyond the mobility vital sun and introduce really looking at chemical, both for security and environmental application anytime, any location. So we rely on electrochemical sensor for three decades for both mobile device and wearable devices we started the 90s lot of wearable for toxic metal and nerve agent. And the last decade we work on wearable wearable for, again, security and environmental application. And again, the best example of success story in terms of in the field is the lesson from diabetic from glucose over the past four decades we see moving from bench top, a large instrument to this disposable strip in the mid 90s and now we have the wearable this is a needle like a cgm. This is the only, only success story the big market multi billion dollar driven by the diabetic. So this is an example of moving over three decades it took to go to the same disposable strip field device to on body sensing. And the advantages of all these are the continuous money we have continuous information about environmental exposure, anytime in location 24 seven. This can be for medical for security for wellness and other application so this is a real time 24 seven data, and we can have a temporal profile of environmental hazard, again, over a long period. And again, this is done if you buy your market we don't need to do blood sample money to pierce the skin and changing the way we do the environmental and else monitoring both for civilian and soldier we do a lot with the military for exposure for danger. These are major advantages, and but the only success story is the glucose so let's see what are the limitation what are the gaps, what are the challenges. You know we are trying to move this big laboratory the standard the classical traditional laboratory to have a lab on the skin. It's a nice vision we would like to have all the lab all the lab operation to do it on the skin under the skin in the mouse on the contact lens, but there are major gap and major challenges in the placing the lab complete lab on the skin and get reliable result reproducible results This is the main question and the major challenges most performance challenges and engineering challenges so we will see there are a variety of challenges related to energy reliability of the data and data safety data security. So all of this we will see the next three minutes, and we go more detail is the performance and let's say a gap in the performance unlike the lab, where we have major challenges, compared to standard laboratory in terms of the scope of the measurement the accuracy, the stability. The ability is a concern because we are working outdoor condition the temperature in Phoenix now is 110 degree this enzyme cannot survive it's not like in room temperature in the lab, changing condition, and then we need to validate everything to be to compare to the gold standard which is the classical for FDA approval we need the building calibration and overall we want to ensure the reliability of the result are not compromised let me use my laser pointer. So we want to make sure that the reliability accuracy are not compromised we talked about. We have a safety safety is more if you do measurement on the skin on the eye under the skin bio compatibility. Many of the essay we do in the lab like immunos they're not reversible and cannot be adapted to on body operation have multiple steps or like a typical Eliza immunos they're in the lab but for a different trace a chemical we cannot do it on the body yet. The issue with the compliance with the body the flexible device you want to match the property of the device with the tissue you need material innovation energy demand was a powering the device the communication wireless communication we have challenges. We talked about big data processing the data data security data safety data privacy and finally the scalability of this device low cost large scale manufacturing along with the regulatory the FDA approval so these are major gaps and major challenges. But we are not giving up and we are trying to address these challenges again in our lab we are working on wearable not only to monitor the environment but different biomarker putting sensor on the skin for sweat monitoring saliva monitoring using mouth guard. Tears monitor with contact lens and micro needle for ISF, but a lot of this is monitored the surrounding environment for environmental and security as all about molecular biomarker we can analyze the sweats and I have a tear looking for metabolite like glucose alcohol electrolyte sodium potassium drug like opiod. Cortisol hormone stress biomarker disease biomarker nutrition again chemical agent toxic metal and many more using our wearable. So I'll give you a few example, a few example from my lab this is for example a glove that we use for swiping looking for explosive residue and nerve agent residue so we have one finger for swiping one for detection connect to a portable device. We have two semis for nerve agent a textile based sensor you see the electronic is flexible as enzyme or pH to selective for the nerve agent or we have micro needle that want to check exposure for opiod or nerve agent one sensor is for the opiod and one for the nerve agent so you can distinguish between this and this. For another example is a ring that we develop for exposure most explosive and nerve agent you see all the electronic is integrated inside the ring. This is the one for the Navy in San Diego for exposure for explosive residue on a wet suit, or this is again for military application for nerve agent on textiles so these are some example. Again we see tremendous progress over the past decade for chemical sensing for all this micro needle mouse gas skin glove, but there is again major gaps that I discussed and we will review. In general, the technique are very promising, but despite of this major progress is still major gap and performance for reliability and scope, but we can address this is multi disciplinary approach people from different field, and hopefully, the field will move very fast, like in the case of a new commercial possibility to accelerate this possibility for on body environmental and security measurement. Thank you very much. Thank you very much Dr Wang we appreciate it. I would like to just remind our audience that we will take all our questions during the Q&A session at the very end. With that I'd like to introduce Dr Ed Ramos Thank you so much for being here briefly Dr Ramos is the chief scientific officer at care evolution and co founder of the digital trial center and director of of digital clinical trials at Scripps Research. Dr Ramos's work is based on the growing need to rethink clinical research studies by leveraging digital health technologies and embracing decentralized sightless approaches which can promote broad participation without sacrificing robust data collection. Dr Ramos's leadership role focuses on overseeing the implementation of digital research studies and a variety of contexts, including infectious disease maternal health sleep medicine and precision nutrition where he serves as the primary investigator. Dr Ramos prioritizes efforts to address health disparities and enhance participation in a diverse equitable and inclusive manner which we know is very important. In his annual role he served in the federal government for nearly 15 years. There he led independent research projects and has been coordinating and administering large scale national research efforts, his expertise spans population genomics by informatics, mobile health and digital clinical trials. In his previous position serving in various capacities at the NIH. Dr Ramos oversaw and managed portfolios that focus on innovative and groundbreaking initiatives in that improving public health. Most recently Dr Ramos was team lead for the participant center of the all of us research program and ambitious program launched by an NIH inviting 1 million people across the US to help build one of the most diverse health databases in history, which could help in the development of better treatments and ways to prevent different diseases. Dr Ramos began his federal service as a legislative fellow and legislative assistant, advising then US Senator Barack Obama on health and science policy. Dr Ramos is a PhD and molecular biotechnology from the University of Washington, where the thesis work carried out at the Fed Hutchinson Cancer Research Center. And with that the floor is all yours, Dr Ramos thank you. Thanks Rima. Appreciate it. Thank you for the invite and congratulations on what's been an amazing workshop to UN team. I have three thoughts. We've had a tremendous amount of discussion around specific technologies, the vision of where these technologies can potentially take us in the environmental space. And as much as I like to geek out with the next person in terms of these different devices I wanted to take a step back and think about how we think about implementing these technologies and what it means to actually bring it into a clinical research study. And considerations in terms of participant enrollment and so on so I'll start with how we typically view potential challenges enrolled blocks in the traditional clinical trial model in which the physical brick and mortar facility is really front and center and that really is what governs a lot of the decision making process around how you structure and design different elements of a clinical trial or clinical research study. And with that potentially brings with it challenges and also limitations in terms of recruitment so for example, if you are a regional academic medical center or a specific provider system. You're typically building your cohort from the surrounding patient populations. And the challenges there is that you may be limited to a certain demographic that may actually pertain to the research questions or may not be representative of the research questions at hand. In addition, we think a lot about what is the burden participants so there are a lot of statistics there around how typically how close clinical trial clinical research sites are to an individual. I believe the average time something around 70% of participants that are accessing a clinical trial, typically need to travel around two hours to get there so you're thinking about now, how do they take time off, how do they access the site and so on and so forth. There's also resource dependent issues in terms of building that study coordinator team ensuring that you have expertise and of course having trained professionals for things like bio sample collection. And a lot of the data that's collected is limited to the time that you have that participant in that brick and mortar facility, which makes it very episodic in terms of the collection. And so COVID-19 really kind of ushered in a new lens that really catalyzed a paradigm shift in terms of how we can potentially move clinical trials out of these brick and mortar facilities and more into people's homes. And this is typically referred to as the decentralization of clinical trials. So a decentralized clinical trial doesn't necessarily have to have a digital component and a digital research study doesn't necessarily have to be decentralized. But it's really the marriage of these two where we start seeing some exciting things in terms of how we can view the construction of these study designs in a different way. So moving from that potentially restricted patient population to really anyone anywhere and opening that up in terms of access, removing the burden of having the participant get to a physical site, but really being able to do it in the comfort of their own home. Leveraging technologies like a smart phone to be able to build a study app and say and provide a self-guided experience. And with that has the positive kind of downstream impact of not necessarily needing to build a huge study coordinator team but really focusing on the participant experience as it's delivered digitally through a smart phone study app. And also leveraging innovations for things that you may not consider as something that can be done in a decentralized way like by the sample collection we have studies that collects saliva and blood and stool again all things that are done in a self-administered way. And then of course most germane to our conversation today in this workshop is talking about all these technologies really give you this ability to not only collect data in a continuous way, but do it in such a way that it's collected in a real world context and that real world data allows for an eventual kind of analysis to get towards and drive towards real world evidence for a variety of different research questions that we may have. And again specific to the environmental space as it relates to this particular workshop. So proof of concept. I'll just kind of leave you with the digital trial center at Scripps Research has been successful in implementing this across a variety of different research modalities. As Rima mentioned, we have interest in the sleep medicine space the precision nutrition space the maternal health space infectious disease. And hopefully you can easily see how these can be adapted for a variety of different research context, especially environmental science and the impact on the public health perspective so appreciate the time and just wanted to give that quick setup and excited to jump into conversation here. Thank you very much. And, and thank you both, Dr. Wang and Dr. Ramos. I'd like to just remind our audience quickly that the Q&A function is below the video player and slide those so please put in your questions there. And with that, I'll take us to our Q&A session. Thank you both speakers and honestly I don't think we'll manage to ask you all the questions that are so interesting coming from our discussions today but maybe to start us off. You know I know Dr. Ramos, you mentioned this and it's very important in your work the issues of diversity, equity, inclusion and accessibility remain front and center. In the space where was her precision environmental health and for medical care delivery. So maybe starting with you and then we'll move to Dr. Wang. What in your opinion, or how are we, how well are we doing in this space in your opinion. Are we being sort of proactive in thinking and designing for diversity, equity, inclusion, accessibility. Are we lacking anything in this vision. What could help accelerate us getting there from any dimensions or examples or sort of lessons learned from your work. Yeah, absolutely I mean as you know, we could have a whole another workshop on this particular topic but it's important to raise and there was some great comments in the previous session with regards to recognizing the need for diverse cohorts. So to answer your question quite bluntly, I don't think we're ever doing a really good job in the health disparities space. I don't know what we could potentially do but where I get excited is that this really is a paradigm shift. So we really are rethinking how we do clinical trials and clinical research. So we're at this beginning stages of a decentralized digital clinical research model that we're still figuring out how to do it and we, and I think we owe it to underrepresented and underserved populations to prioritize them. And to say, you know, I will figure this out in the majority populations first, and then we'll get back to you and figure it out. And the problem there is that we start developing an infrastructure that is suitable for only a subset of the representative population. And by the time you get to other populations and other communities, there needs to be additional activation energy and additional efforts and resources to put forward addressing some of the very specific contexts that they bring within their communities. And it's like recognizing that not everyone has an Apple watch that there are gaps in service with regards to bandwidth issues and rural communities that there may be trust issues, specific to certain communities around their sharing of digital health data that we have to examine and what our security principles are with regards to how we not only ingest data but also how do we secure it as a research team. In addition to recognizing that we should be figuring out how to communicate what are these, what these new technologies are and I think there's, I've had a lot of conversations in which we can easily slip into an agist kind of framework where we say, oh, you know, the older people know how to do this. So we're going to set them aside. And I think some of the success that we've had is that we've shown and that that's just not true. I always bring my mom up who's, you know, getting close to 80 years old. And she's been my number one participant in darn near all of my studies. And I've been pleasantly surprised at really her ability to navigate and it's been a tremendous amount of feedback from her to say this isn't clear. This isn't right or you have to think about finger strength for this particular device. So a variety of different things. And again, I can go on and on in this particular topic area, but it's, I'll end by reiterating this is our opportunity to prioritize a lot of populations and communities that we typically skip over and wait for the second round. The second round to come around. And so I think we have an opportunity to really make that our primary loans. Thank you so much. And Dr Wang, I'd love to also hear your thoughts on this issue and perhaps from your, you know, technology and global perspective as well. I think I'd cover it nicely. I mean, we developing the technology. The key is that the personnel in my lab are very diverse. My student postdoc who will be in the field are all from very diverse population here in the UC. And again, some of the luxury item, you know, if you do wearable for wellness or nutrition, this is more depend on the, on the, how rich you are, but most of the one for protection of the civilian or soldier. Here there is no discrimination between different, different populations. So it's not like it. As you mentioned the apple watch people use it for wellness, recreation, personal nutrition. This we need to improve that it will be available to. But when it comes to protection against environmental hazard or nerve agent, this should be a general available to all the population. And on that note, perhaps I'll ask one more of my questions and then we'll make sure to get the audience questions in as well. So you mentioned, you know, your team is highly interdisciplinary and I think from all the examples we've seen today, this is such a highly transdisciplinary field and you need so much expertise as Dr. And Sally also mentioned before. And as you mentioned, perhaps like the teams at play and running these studies now are shifting and changing and so on. So at least in my experience, I still feel like it's very challenging or the systems we work within are not very well built for interdisciplinary true collaboration and I'd love to hear both your thoughts. And what is it that we need to kind of help us break through these challenges and really work in highly transdisciplinary, you know, functioning environments. Perhaps Dr. Wang, if you'd like to start. You know, in UCSD we have a fantastic multidisciplinary we work closely with the physician and, and again the old environment is for equity and diversity, both in terms of training and accessibility. So we see no major barrier I mean the barrier is when people need to get expensive gadgets like Apple Watch and other many for wellness and nutrition but not for medical or security application. Yeah, so I'll add a perspective slightly different although I agree with, with all of these points I think the workforce is extremely important to recognize diversity in thought, as well as in background. But for us, we've landed on one of the greatest assets in developing our research efforts and that's putting together our virtual advisory team. So not interdisciplinary in the sense of professional backgrounds and disciplines, whether it's engineers or technologists or computer sciences, but really interdisciplinary if you think about the different roles that people have in our lives so patient advocates and industry and non profit organizers, grassroots organizers, the stay at home mom. So that's, that's our interdisciplinary lens in terms of ensuring that we get those perspectives because our expertise really is in the implementation science, and we can't be experts in that implementation science if we don't have a recognition of what the risks are. Ultimately, this has to be useful at the human general population level. When we talk about public health and so we really try to kind of reverse engineer that and Sarah well who's actually using it, and how can we extract a lot of those insights interdisciplinary, transdisciplinary in terms of professional disciplines 100%. We need technologists, we need clinicians, we need experts in public health, we need regulatory experts, we need all of that. What I'm just suggesting is that additional layer of what is the diversity and thought in terms of discipline with regards to the general public. It's very interesting. Thank you and perhaps I'll take a question from the audience. This is directed directed sorry to Dr Wang. What is it at the future of using tattoos or lab on the skin for monitoring analytes like glucose or environmental factors. And again glucose is the best example has been a, we all learned from glucose from three decades of glucose remember started the 80s people were dreaming about implantable glucose. So it took two decades until we realize this CGM continuous glucose, but this is driven by the market I don't see the environmental market is big enough. But the same we can use the same technology for other metabolite electrolyte, but it's each of these need the investment and I bet there is a big company like here Dexcom or about the investing billion because of the market so environmental market is relatively limited The US government for security applications investing a lot for protecting the soldier but I doubt that just the regular environmental will be major investment, unless there is some requirement. But what was the second question, it's mainly I think, I think you addressed it it's basically also for environmental factors let's say similar types of sensing. There is a really limited mind the technology is available and we need somebody to invest in it to make I mean, even the glucose company like apple is dreaming to have a glucose on the watch and the investing million is still drawn. So everything is challenging especially for the chemical sensing. So the chemical sensing is a different challenge and mobility and vital sign really the stability of this bio receptor the end of the end, they are not so stable for prolonged operation outdoor extreme condition in the desert and so on. Thank you and I believe being a mention that as well. We have a question specifically for ads. What is the type of information that you are collecting with decentralized trials and is it self reported behavioral information. How confident are you have the accuracy of responses when there is no coordinator interviewing these respondents. That's a great question and dovetails a little bit with the previous one in terms of we can have specific sensing for environmental measures. But it's also important to recognize that we can layer on top of that other digital biomarkers that are still helpful to kind of understand the context, whether it's measuring physical activity or sleep duration or resting heart rate or respiratory rate. All of these things I think could nicely complement more environmental specific sense and metrics that that we could glean and so those are examples to the initial part of that question in terms of different types of data we collect, you know, things that you would normally have a wearable from a Fitbit or Garmin or whatever. We have the opportunity to use a study after deploy a number of different surveys and so with regards to participant reported outcomes that can be in the context of validated survey instruments and we can get into things like anxiety disorder assessments or depressive disorder assessments stress and mood. All of these things that have validated instruments that have been deployed in the field in traditional studies. The last part is the trickiest one with regards to how how much can we really put stock into the accuracy of the of the answers that are being delivered. So a couple of things. One, I think this is probably most a question that surface the most in a regulatory context. Right. So when we're thinking about shifting this from a traditional research study to tradition to a traditional and formal clinical model, you now have the lens of the FDA on top of that where they're asking. The initial model was you're bringing 1000 people into one spot and one person's collecting the blood pressure. Now you're telling me that you're giving me 1000 blood pressure readings that were taken by 1000 different people. What does that offer up in terms of accuracy and validity. And those are things that we just have to think through and those are things that our team thinks through and this is what we get excited about where are the challenges how far can we push this and how much can we really determine a signal above the noise of potential inaccurate readings or readings that wouldn't necessarily be valid. And with that kind of blends the strategy of in ideal situations in which you can mimic the gold standard collection of a metric. Yes, let's bring that into the workflow but in instances in which you can. Does a silver or bronze standard metric serve the appropriate purpose and so for something like steps. I wouldn't put any stock into the exact number of steps, but I would put stock into the delta of steps so it may not be 10,000 but when you have recorded and create a baseline when an individual 10,000 steps. Now all of a sudden they're walking 6000 steps or 6000 steps are being recorded. It's really that delta that we get most interested in that we can do quite a bit with in terms of profile analysis trend analysis. And potentially even predictive analysis so it's a great question. I wish I could say I figured it out completely but it's something that we certainly are a front brain of all the time. It's very interesting thank you think the methodological sort of questions and ways we approach things are really fascinating now and it's sort of a new world of how we think about doing these studies. The next question is from cloud you and our planning committee and for both our speakers, maybe starting with Dr wrong in a futuristic world where we are all wearing sensors providing us with information on real time environmental exposures and health indicators. What do you think will be the balance between desensitization from data overload and changes in the use of consumer products available for everyday use in the marketplace. Meaning will people sort of reach a point how do you reach a point where people are not just overloaded with information and they're actually moving towards changes in their behaviors in a useful way. I mean there are different purpose one purpose is to give you warning and alert in case of any exposure sudden exposure that you need to take some precaution. So this is different than have the continuous data temporal a profile of this environmental pollution in San Diego, but the main concern when there is a real sudden exposure and the warning if you ever. As we say in the military for Norwegian explosive but the same for a civilian. So I don't think that the normal population worry about the long term temporal change as long as I was in the limit of normal and not exceeding levels so the main goal of this is to give an alert and warning for sudden sudden exposure from my point of view. But another area that we think about quite a bit in terms of behavior of the participant and I think there is quite a bit of desensitization that we have to contend with because there is a significant amount of data overload that we get on a daily basis. So where we land and this allows me to kind of bring up another important aspect that we focus on is return of information return of data return of insights back to the participant. So if you start including them into this process of don't just give me all your data but let me actually provide something back to you I think that starts not only establishes trust and I think a meaningful relationship with the participant. But also gets them engaged and motivated and understanding I think one of our successes. And the glucose keeps coming up because it's a great exemplar, but we have a precision nutrition study in which we launched and we provided a DEXCOM G six. But we really needed to incentivize and motivate the participant to log their food to where they're fit fit and do all of these other things that we were asking of them. In addition to the monetary incentive financial incentive which can help it really ultimately what motivate them was showing them the real time continuous glucose monitor readings. Overlaid with the macro nutritional information that they were recording from the food so that they can very intimately because it's personal to them see what this data means. And they actually were starting to understand the whole purpose of the study I think back at some of my older studies that I used to run. And it was very much fly in fly out and the participant really didn't their head was left spinning in terms of what exactly that I actually just participate in. But being able to be a part of, oh, this is my information and you're providing it back to me in a way that I can understand it and making meaningful. I think that's what can start chipping away at the desensitization because it starts really bringing this element of being meaningful for the participant. It's really really challenging and really difficult. And I will say thank you so much because that was going to be my next question and you're already answered it, especially about returning information and getting to actionable results in an ethical sort of, you know, useful context sensitive way. We are at time for our Q&A session and as I predicted, you know, I wish we could go on for another two hours. But I'd like to really thank Dr Wang and Dr Ramos for your time and for sharing your insights with us. This has been tremendous. For now, I'd like to move us all into closing remarks and thank you again to our two wonderful speakers. Thank you very much. Thank you very much. Thank you. So, I hope everyone is still online with us. And this brings us to the last day and the last session and closing of our wearables workshop. I would like to express my sincerest gratitude to all our esteemed speakers who have shared their expertise and contributed to these discussions that have unfolded throughout the day. Your valuable insights and active engagement are really shaping this innovative field of wearable technologies for precision environmental health and for biomedical research, leaving a lasting and meaningful impression on our collective journey together. And I'm sure everybody online and watching us today feels the same. The first day of the wearables workshop covered a range of topics including non-invasive monitoring techniques and the development of wearable devices. We explored discussions on improving personalized health tracking, monitoring indoor office design, which we learned of course as an area we spend a lot of our time in, over 90% of our days. We learned about utilizing silicone wristbands for personal exposure detection, particularly for vulnerable populations, about wearables for air pollution monitoring and extreme heat, about monitoring warfighters exposure to chemical hazards and so much more. And today on our second day, we just heard from leaders and innovators in interdisciplinary fields and the biomedical space. The session three speakers highlighted a variety of wearables and sensors to detect and collect physiological biochemical health and environmental data. Data derived from wearables and other health data sources can then be used to train and develop machine learning algorithms for disease predictive models for diagnostics and for treatment. As a deeper dive, we heard from Dr. Mahalinga and Dr. Chang about the women's health study with Harvard School of Public Health and Apple, which was used to track fertility and even exposure to wildfire smoke in women. Dr. Jesselyn Dunn's presentation focused on digital biomarkers and utilizing smartphone collected data transformed into predictive health indicators. So Dr. Dunn's use of machine learning classifiers, wearable measurements and case studies further demonstrated the power of digital biomarkers in cardiovascular disease and diabetes monitoring. We also heard from Dr. David Armstrong, who highlighted the significance of wearables and managing diabetic food complications, a common and often neglected issue with frequent amputations occurring every 20 seconds. Dr. Armstrong touched on the importance of providing patients with devices that compensate for their lack of pain sensation, such as smart boots and exploring innovative sensors and approaches, including smart socks, smart shoes, insoles and injectables to positively impact patient outcomes. And then we heard from Dr. Vina Misra on the assist program that highlighted the advantages of long term health monitoring and the need to understand the relationship between health and environmental toxins through innovative personalized wearable technologies. And then we just had our engaging panel discussion on understanding how technology adoption, implementation, report back and science communication factor into advancing biomedical and environmental health research. So, excuse me, I'm talking about session four, not our session five speakers just now, but session four speakers highlighted areas concerning data privacy, cognitive liberty, data equity, workforce development, community engagement, standards and regulations and so much more. So just to give you a recap with a little bit more granularity, Dr. Nita Farahani's presentation explored the convergence of wearables and brain technology, highlighting the embedding of brain sensors into everyday devices and the promising potentials of wearables or newer tech and the considerations around cognitive liberty. Dr. Ritika Chativerdi discussed health data for equitable precision health. Dr. Chativerdi's research focuses on biases and wearable studies emphasizing the need for socio-demographic representation for developing benchmark data sets and adherence to fair data standards and principles in AI and machine learning research. You then heard from Dr. Shikhar Bansali on the use of wearables in public health and the challenges of interdisciplinary training, highlighting the need to update and enhance educational curricula to effectively address these areas and attract talented individuals and students and trainees, but also discussing the influence of social media and data-driven hype. Dr. Tiffany Powell Wiley discussed her community-engaged research approaches in several Washington DC communities at high risk for cardiovascular diseases. She is using wearable technology to discover how neighborhoods and other environments influence the development of obesity, diabetes and other markers of cardiometabolic risk. And then we heard from Ms. Deborah Prince about the importance and the reasons why UL developed standards for the safety of virtual reality, augmented reality and other wearable technologies, highlighting the process for how to develop appropriate standards. And lastly, moving on to our session five group discussion that we just had on the future of wearable technologies. We discussed the current state of wearables, future approaches and gaps and challenges we expect in getting there to the promise of wearables for all these applications. We had Dr. Joseph Wang discuss innovative approaches to wearables such as gloves for chemical detection, innovative data collection and so much more. And we also had Dr. Ed Ramos during his lightning talk discuss a future beyond clinical trials, one that is decentralized and really focused on the patient and participant experience. So over the course of these two days, we have witnessed the conversion of brilliant minds and groundbreaking ideas in the realm of wearable technologies. As we bring this wearables workshop to a close, we do though with a we do so with a deep appreciation for the knowledge we gained for the perspectives shared for the collective commitment to advancing this field. I would like to extend a huge thanks to everyone involved in making this workshop such a huge success. And we encourage you all to join the standing committee on the use of emergence emerging sciences for environmental decision, June 14 through June 15 for our sister workshop on advances in multimodal artificial intelligence to enhance environmental and biomedical data integration. And also personally, I'd like to thank our amazing planning committee and program managers and leaders in getting us here. And with that, we hope you have a great rest of your day. And this concludes our final day two of the National Academy of Sciences wearables workshop. Thank you, everyone.