 Hello, everyone, and thank you for coming. My name is Robert. I am the data science program chair of the U.S. and we have organized the data science summer school in the U.S. for two reasons. And I would like to talk about to have a course, a crutch course on data science and healthcare. And we also kindly agreed to have this open talk about how the sense can be used in healthcare. Now, please start the interview. Thank you. Thank you very much, Hadek. Thank you all for coming here. I'm really honored to see so many people here today. It's always a pleasure for me to be back in Armenia and to meet people in the space where I work. So many of you here. I met some of you. I met previously. But I will briefly introduce myself. So I have graduated here. I have graduated here in Yerevan State University. Now I live in Ireland with my wife and three kids in Dublin. I did my PhD here in the National Academy of Sciences. I started my time in the University of Cambridge in the center of medical life statistics. Then I worked as a GIS, a management information system officer here in Armenia working in different organizations, mostly funded by the UJI. Then, since 2008, I moved to Ireland where I was invited initially to work as a post-doctoral researcher. And then I got my degree scholarship. I moved to the United States where I worked in the University of Ireland for a few years. Came back to Ireland and now I work as a data scientist about a year ago. I started to work in a company called United Health Group including the data science team there. So, yeah, you can see the picture of the building where I work. It's very beautiful. This is where I am. I'm happy to talk about data science and big data applications and emerging technologies in healthcare in medicine. So, this is an interesting picture. And when we talk about healthcare, we need to realize that actually we are talking about sick care because our systems are set up to treat sick people. They are not really set up to keep us healthy. And very often we end up in the hospital's plan but this is already there. We are waiting just to get the diagnosis. And often, fortunately, it's too late. We have this continuous data in healthcare. We are having our daily visits to our physician. We are getting our blood pressure checkups. We are getting some vaccinations for the blood vessels or getting some EKG. And this is just this continuous data. We are not getting a constant stream of data in real time which could help to get better decision about our health and about ourselves. And it is not surprising if you can see that the format used currently in many countries and in many hospitals, so there is a huge amount of data stored in paper format currently. Although there are many companies that are working to digitize this information, there are many hospitals who are using digital information. And I actually met today Garen who is working on digitizing information from Armenia, so we are happy to learn about it. But still, there are a lot of hospitals using paper format, using flux machines to transfer data. And it's not effective. And it costs us time. It costs us money. And this is a picture from Hungarian hospital where they monitor the patients, they track the patients using this manual traffic system. You can see the patient index card is there. So this is a picture which you would expect from 20th century hospital. You wouldn't expect anything to see like this in a modern hospital. And this is something where digital technologies and big data technologies can help to improve and make more effective product. Otherwise, we are spending too much time. We are spending an hour waiting our doctor and we are getting just about 10 minutes, 15 minute visit afterwards. And when we are having the visit, often we don't even remember many details what was wrong with us a week ago, why we were not happy, what was the specific pain we were feeling in which part. So the doctor is asking a lot of questions but we are not being able to answer any questions after the meeting. And when you consider how difficult it is for the doctor, this is another issue. We have about 15 minutes with the doctor and the doctor needs to understand what's wrong with us, he needs to ask a lot of questions, he needs to do physical examination and understand what's wrong with us. And then he needs to recall the research down in that field and based on all that knowledge hopefully he will be able to give us different correct diagnosis and treatment. And of course it's not easy considering the amount of information that the doctor needs to do correct diagnosis. And that's why many physicians are under stress, many patients are not happy and the survey results shows that about 60% of physicians are feeling that they don't have enough time to spend with their patients and many patients, about 63% of patients forget to add their medications and there are many people who have chronic disease, so the situation is not good and this is where electronic health records come to a system. Electronic health records usually are created as a digital alternative for data records. So there are digitized versions of this data we are collecting during our medical visits, our hospital admissions, some lab tests and different lab results and it collects and stores and patients have information basically. There is a huge industry there so you can see the numbers in 2016 it was estimated 23 billion live of all HR and just after a few years 10 billion is estimated to grow. So there is a huge potential there and many companies, many countries are actually trying to reward in this sector and it's necessary because there is a new direction for patient care now. Until recently there was a big disconnection between patient-generated data and hospital-generated data. We know that there are a lot of verbal devices we all have mobile phones we have different sensors and mobile devices and we collect a lot of information about ourselves our activities, our environment unfortunately this information is not shared with our clinicians, with our health organizations and from the other side our health organizations are collecting a lot of data when we are visiting the hospital we are doing some medical images so there is a lot of data collected and stored in organizations and there was no proper connection until recently. But now with the technological progress we have this opportunity to synchronize these data so we can actually share the data with health organizations and put the data back this will give us a better understanding where we are now and it will help us to be on the same page of power with doctors so there will be not just daily tasks there will be not just daily tasks but the doctors will be able to see the real-time data and understand something is wrong going on with one patient I need to call to have another visit tomorrow instead of waiting a month or identifying the patients who are at risk for some particular disease or so so there is a lot of potential there to share this kind of information and there is a huge amount of information coming from different sources so in this chart you can see some of the sources particularly used in healthcare and clinical data is on the small portion here as you can notice some of the data used in medicine like x-rays MRI, CT scans is another portion of this data you can notice that genome is quite a big portion of data covering a big portion of data there are a lot of other data coming from metabolome, microbiome device data again we use these other devices we use these activity trackers or mobile devices so they generate a lot of data in healthcare organizations we are exposed to pollution and environmental hazards so environmental information is also very important for healthcare this is also something we need to consider and finally the social data our personal circumstances our geographical location our neighborhood these are all affecting on our health so just considering clinical data is not enough to get to be healthy basically to use all information we can and the relationship between physician and client patient is changing nowadays we are more educated, we have access to more data, more research we understand we do some research ourselves and we want to take part in decision making we don't want just to upset what was thought, we want to change this decision this is changing and mostly this is changing because of technological advances we can see that the access to internet worldwide is rising exponentially so there is a huge amount of people now have access to internet quite about let's say 10 years ago the number was much, much less and it's not just internet if we consider total population in 2018, 7.5 million over half of them already have internet access about 68% already use worldwide funds and there is a lot of people using social media and other internet activities so this all helps us to generate data and use data to be data users and data providers and this is mostly driven by Moore's law which is basically showing that the computer capacity is doubling every other year this is why this is why our mobile phones now are like super computers of the 1980s they are more powerful than the computers that they are using in NASA when they were running on the moon this is why we can use this mobile devices as a medical device we can use them as a mobile medical device to hold our health and when we think about our future we do start to think exponentially otherwise we are doing bad predictions if you take a look at this chart like this and it is growing up to the present days there are several ways you can do the forecast for the future you can take into account the trajectory based on the past project then it will be very bad prediction but here we can take another approach taking trajectory based on the present growth rate instead of past growth rate it will be a little better but again to the bad prediction the correct prediction will be if you take into account exponential growth and to make this a little bit simple this is a nice chart that I borrowed from Wait By Why website that shows the surprise factor for exponential growth so very often we think that our life is very predictable but in fact if you think exponentially you never know after 5 years what kind of technologies will be here which we are not even thinking about now and this is not just one technology this is combination of multiple technologies we are seeing advanced scheme all different sectors like artificial intelligence surgical robotics and remote access and so many things are developing and innovation comes when these different technologies are there together when they are combined and we are seeing this kind of innovations now we are already used to this digital payment types we are changing our way of watching movies we are using different type of electronic books instead of general books we are changing our behavior when we are booking flights or booking taxis now and this is continuing this is changing and this is so called a modification effect transforming everything starting from finances ending in healthcare industry and company like Uber hasn't invented anything new they just combined invented online maps they haven't invented online payment systems or GPS device they just layered all these technologies together and after 3 years the value of the company grows by 18 billion and this is also this is also driven by demographic shift now we have growing number of population millennial population millennial generation our digital natives they are growing with these technologies they are it's very important for them to use these technologies and they want to get these kind of technologies in their life and when we are asking what is the idea of healthcare experience for these consumers the convenient access is the biggest one the survey done in US show that 59% wanted to digital healthcare experience and they wanted to mirror retail actually people are quite happy about retail experience and they want to move for healthcare to move to that direction 74% of millennial patients buy the ability to book a point and buy it and pay online and 48% want to partner with healthcare provided for personal lives so there is a big shift and this is very attractive market also it's not just it's attractive for all companies both in healthcare and in technology because there is a lot of money like the forecast for 2025 healthcare landscape and your revenue is 5.5 trillion US dollars this is very attractive and we have giants both in healthcare and in the tech industry tech industry have huge experience in technical side but they don't have much experience in healthcare so they are trying to shift there from that side why in healthcare industry we have companies who have great potential and great data sets covering healthcare and medicine but they don't have enough technical capacity so they are trying to increase their technical capacity so everybody is trying to get a piece of this nice 5.5 trillion and there are a few nice examples we all heard about Ion Watson which first became famous when it won the game in the GeoParty winning all time two winners, two champions of the game and another example is AlphaGo from Google and you may remember the headlines when AlphaGo beats the gold master Lee Sedol who was all time best gold player and considered that the computer would play gold in such a level so both IDM and Google are moving to the healthcare sector also IDM Watson was used in German hospitals trying to do this diagnosis they were using a lot of data provided from different sources and then patient information and the idea was that Watson would give the diagnosis and the diagnosis will be checked and then it is collected of course it was not so simple IDM was one of the first in the sector so it was not totally successful many people working in hospitals but it was failed because Watson wasn't able to diagnose even some simple diagnosis for example simple cases when there was a pain chain chest chain or high temperature diagnosis of flu Watson could go into some rare diseases of cancer blood cancer and so on so it was not very satisfactory the other application was Google AlphaGo they are using it currently in the United Kingdom with IEC's treatment and they are training a lot of models using OCD scan of eye images using AlphaGo for diagnosis however again there are some issues there there is privacy issues many people in the UK feel that there was a privacy breach for data so there are a lot of challenges there about companies but there is a big investment in big data so we can see the numbers in 2018 there were about 4.7 billion estimated value after a few years to go there growth in data is also a phenomenon healthcare data is growing hugely you can see 1013 volume estimated 153 exabytes in 2020 it is estimated to grow up to 2314 exabytes just to give you a perspective this is a huge amount of data we may not even be able to store this data so if we take the results from cisco global cloud index report so you can see that by 2020 it is estimated that we will have to pass for 985 exabytes which is 2.5 times less than just data will be generated by medical devices something healthcare in general so there is a huge challenge even to store data you can imagine how difficult it will be to use this data we will need a lot of machine learning and just computing power to analyze this data to make it useful and if you adopt this big data technologies this will make it the best so you will see the results here from the report which says that there will be about 20-30% in cost saving there will be about 35% rising patient access and 20% improvement in outcomes and about 30% growth in revenue so this is very attractive and what would be the verification of healthcare look like same verification here this is an example so here is what will look like healthcare to combine these different technologies and make them user-friendly make them accessible and make them easy to use so we will go through some examples of different applications and medical devices just to give you an idea what is already applied and what is actually coming soon so this is a price cost for general sequencing so you can see how much it is reducing the price so it went down significantly already it is couple of thousand US dollars you can do the genome sequencing but why this is important of course we don't need our genome to know our genome we need to have the input and we need to exercise but there are some diseases which we need to do genome sequencing to understand we are prone to have that disease if we have that gene which may cause the disease in our future and this become very popular with Angelina Jolie when she announced in 2013 that she carried the gene and the gene put her under significant danger of having breast cancer so she did double mastectomy to reduce her breast cancer risk and this is called Angelina Jolie effect now many women are considering this example and they are thinking to do the same some of them are doing of course there are many critiques of this approach saying that ok having cancer as gene doesn't matter it doesn't mean that you are going to have cancer always so to have such kind of extreme intervention maybe not necessary but again this is information we can have and we can decide if she wants to do that or not and it's not just genome or medical data it's also environmental data so there is a big effect of environmental pollution on our health this is the report from UHO World Health Organization so 23% of all global deaths to our environment this is roughly 12.6 million deaths a year and you can imagine how big is this because these are lives and the UN report also shows that particularly air pollution is a big it has a big effect on our health and about 3.7 million deaths attributed to air pollution on it it is well known the effect of air pollution on particular diseases for example it is known that asthma air pollution is a big trigger of asthma many people are suffering from asthma there are over 230 million people worldwide suffering from asthma and asthma attack can be vital people are dying because of it so having a device, having a data which can help to reduce this number is significant that's why there are some technologies developing projects are working on this kind of problems and I wanted to bring an example in Air News project which is about the small city in Kentucky in the United States Newsweek is a tough place to live for people with asthma or any respiratory diseases because the air pollution there is very high so asthma is a big problem there and there is a company there who started to use smart asthma inhalers to geolocate to map the risk areas of asthma and see how they can improve the situation so you can see the map of all collected data points so 1.2 million data points were collected using these smart inhalers and then another 5.4 million environmental data points were collected from environmental stations so this information helped to understand how asthma attacks are related to environmental pollution it was understood that when air quality gets worse people are using inhalers more often and when the temperature is hotter people are using it often so this sounds natural but these are the facts and in addition they have done the mapping of asthma so you can see their hotspot analysis they have done the risk map of asthma these are the hotspots where people are suffering and using asthma inhalers in the city and then they have mapped the streets they have mapped the trees and they show that ok they have thick streets where it is polluted and these are dangerous areas for these people and they have green areas these are nice areas for these people because there are a lot of parks there are a lot of trees there so these kind of analysis and they have provided an app which actually could do a warning if there is air pollution in particular part of the city so people with asthma would know that ok I better not go there because it's very polluted now this time not yesterday, not week ago but now it's real time and this project was quite successful it was about 82% reduction in asthma rescuing other use and rescues when they had this asthma attack there was 29% of uncontrolled participants gain control of their asthma on average participants more than double their symptoms in a few days so we are talking about people we are talking about their quality of life so this is significant we all know the damage of physical inactivity we all know that we need to be active we need to exercise but small technologies like verbal devices like the Fitbit device which I am wearing now so this gives us some statistics this gives us some simple numbers about our daily activities and this motivates us we are looking at it and we know how many steps I have done today how many steps I have done yesterday how many I need to do to be healthy how many hours I have slept so these kind of simple things can have big influence on our health this is not a big deal I have slept but these kind of small devices helped me there are a lot of single devices devices which you can wear on your clothes devices you can wear on your feet you can have in your glasses in your eye lenses of course if you are taking the survey results not everybody are happy to have devices with their contact lenses but 42% would die there on their wrist and so on so there is not a possibility to have these kind of devices there are more than 40,000 applications related applications in Google store or Apple store of course most of them are not very useful but for basic things like tracking our walking distance tracking our sleep these kind of apps can be very useful and motivating and the biggest factor in healthcare is estimated to come from marketing or IoT so these are the small devices which are connected to internet and they generate a lot of data so there is a huge marketing global marketing is growing usually here so you can see the phenomenal growth anticipated from 2017 to 2025 from 120 billion to 530 billion this is phenomenal of course attention there taking into consideration aging population taking into consideration growing chronic diseases specializing in healthcare courses these kind of technologies may help us to reduce health costs to improve quality of life and make the better place to live another big sector is mobile health so mobile health technologies is also anticipated to grow significantly from 4 billion 2016 to 11 billion 2025 so that's why not surprising there are so many apps now developing the health sector so many mobile devices are developing this sector because there is a huge potential there and now you can use mobile devices already to do blood pressure monitoring you can do monitoring your health rates or performing ultrasound scans or analyzing the endurance samples so these kind of technologies already exist you can even have smart toothbrush which will give you an idea about your brushing habits you can have a smart scale which will tweet about your weight actually it's a good idea but you can have it you can have a device which will show how much you have drink today are you allowed to drive and these are all already existing and Google has a project called RR developing a phone which is composed from several different components and you can have a component which is designed for checking blood sugar for diabetics for example you could have some sensors there which is checking air pollution for asthma and so on so there is a lot of potential there and we are getting into our young citizens also I think you probably don't need to understand about this smart hackers are doing they are kind of tweeting whenever there is some information coming from all the youngsters so sometimes maybe there is too much information there and this information can be stored in different places this information there are already technologies to store this information so you may get tattooed after hospital visit for example and there is a big research going on in gestibles so wearables are becoming in gestibles we are swallowing smart devices and this is very big this can replace endoscopy for example you can swallow a small camera and you don't need to do endoscopy because smart pills are also very important technologies like future mirrors may be very interesting you want to understand what you look like if you eat something or if you don't eat something you may want to know how your skin will look if you smoke before smoke or after smoke you may want to know what your face will look like if you use too much Facebook and these digital devices are becoming mobile doctors need to you need basically to visit doctor to get these checkups and doctor has this equipment in his office or in hospital but now there are a lot of mobile devices you can have a suitcase with a lot of devices and doctor can actually walk to the patient if it is a remote area or you can even have the device yourself you can stick the device into your mobile phone like in this picture here you can do the 3D examination there is not a similar small mobile devices that are available now like in this example you can see mobile devices measuring blood pressure temperature, oxygen level breathing rate, doing AKG so there is a lot of things going on so these devices already exist these devices are providing a lot of data there are some special devices designed for people with epilepsy so this device may help to predict when the person may have epilepsy episodes so you can prevent it or at least you can get some assistance there is this device which helps to do urine analysis of phones so very easy to do the analysis and send it to a doctor and then if you need you can go for a treatment or just get a diagnosis this is a device which you can connect to your phone and get your blood pressure checkup very helpful for people with diabetics the results can go to the doctor immediately and this all brings data this is all composed of big data we are getting out from this all different sectors and now let's imagine we are getting a lot of data we are quite healthy but again we need to visit doctors and this is where telemedicine is coming so online consulting is already very popular in many countries you can get 20 per hour service you can get service from very remotely positioned position very well known position physically will be not able to attend so there is a lot of technologies going into telemedicine and there is a lot of technologies in artificial intelligence in robotics and they are used even for predictive mental health outcomes like you can see here the ongoing reports better than doctors of predictive mental health outcomes naturally that is true but at least there is a work going on in the next 5-10 years we don't know where we will be with this 3D printing is another big area it has a phenomenal growth in recent years in 2010 only 3 hospitals in the US were using 3D printing in 2016 already 99 hospitals are using 3D printing 3D printing is used in different sectors for healthcare you can use it to print artificial bonds you can use to print prosthetics racist supports even tiny organs poly pills are pictured here these are 3D printed pills for patients who need to get multiple drugs so they can combine everything with one small pill so these are also very popular and of course you can use 3D printing when you have a fracture you have broken bone so you can do your x-ray you can do your 3D scum of your hand for example and then you can get customized support for your hand robotic arm like u-arm is what is reality if you take a look at the status you need to watch the denominator and you can just go to the view card and by 20 years it's under research, under development but very soon you will have this too virtual reality is also a big in this sector originally it was always connected with entertainment but it has a huge applications in healthcare it is used to train healthcare workers particularly for specific surgeries which are very expensive and very difficult but you can use virtual reality to train the surgeons and you can use virtual reality to treat mental health problems for patients so there is a phenomenal growth you can see the growth in 2017 it was 8.9 million while in 2022 we are expected to have 285 million regularly and robot assisted surgery is another big application there are a lot of advantages in robotic surgery it is minimally invasive it is more precise in control that's why if you take the numbers there we have 36% before and there are 65% after learning about the benefits the applications and this is becoming something like our cars now the modern car has over 200 sensors which enforce the driver we see all these warning lights engine lights and we know you need to take our car imagine that you are having similar kind of system you are getting a lot of information from different sensors on your body on your clothes and your phone and this data goes to the cloud your doctor may see the data or the artificial intelligence may identify the risks for you and you will get a message by saying ok you better not go to the restaurant the food there is not healthy for you today doctor because your heart seems to be a strange thing so this is already happening and we are all used to information sharing when we are dealing with google map or air traffic for example we can see how cars are driving in the streets and we are dealing some data and in return we are getting some information we are deciding which road to use where to go where is no traffic similar air traffic data exchange is going on there and imagine that you can do the same in healthcare because we live in social communities if we share the data we are making our social community more aware about general environment about what is happening in social of course there are some privacy issues but even some anonymized data you can improve the situation you can get better educated about your environment and your social network and if you know your social network you may decide one day do you need to shake that hand or not because they make danger there and there is a huge advances in technology just this graph shows that in 1913 the average life duration was 34 years just after about 100 years it is more than doubled so there is a huge huge impact of technology and research for healthcare and this is continuing so to summarize I think we are living in a very interesting era where we have all these technologies emerging and we need to realize that medicine and healthcare is changing from this continuous to continuous from reactive to proactive and very often technology is there before we realize so we need to use it we need to apply it we need to create our own innovation thank you