 Well, good afternoon everybody everybody who's in the room and everybody who is watching us by our webcast My name is Tos Cochran. I'm now the director of neuroethics at the Center for Bioethics here And I want to welcome you to the second in our neuroethics seminar series which is a really Exciting venture for me in particular, but hopefully for you as well one of the ways I know that Neuroethics is exciting and part of the future is that I subscribe to a science fiction magazine called as a mobs And every every other story is about a neuroethics topic about artificial intelligence and about Living in a computer simulation and neuro enhancement And so I know that this is a topic for the future and I and tonight's topic is no exception before I introduce our speakers and Say a couple words about the topic I want to give Bob Trug who's the director of the Center for Bioethics a chance to welcome you all Initiative and very much Thank our speakers for taking the time to prepare the material and come down and Anyway, thanks for everyone for being here So we've got three really terrific speakers here to talk to us about brain-based lie detection and their use in the courtrooms although we won't necessarily Chastise them if they talk about the implications of live lie detection Technologies outside of the courtrooms We've got three speakers who are going to talk about various aspects of the topic Our first speaker is going to be Kate Pivo Verova who is a researcher and Transitioning into her role as assistant professor in the law and psychiatry division at UMass medical school And in the Department of Psychiatry She's a former fellow at the MGH Center for Law Brain and Behavior and in the Harvard University's Department of Psychology She completed a clinical forensic fellowship at the University of Massachusetts medical school her research is at the covers a wide range of topics including malingering and deception application of neuroscience to the courtroom Juvenile justice issues judicial understanding of eyewitness testimony problems clinical decision-making about competency Etc. And so she's she's going to be talking to us mostly about the science that's involved here Francis Shen will be our second speaker He is a professor of law at the University of Minnesota where he teaches and directs the Shen Neuro law lab And he serves as the executive director of education and outreach for the MacArthur Foundation research network on law and neuroscience And currently he is a visiting scholar at the Petrie Flom Center for Health Law Policy Biotechnology and bioethics at Harvard Law School, which is one of the main reasons We get the advantage of his presence tonight, so we're very lucky there Dr. Shen does research at the intersection of law and neuroscience and recently co-authored the first law and neuroscience textbook And has published on a wide range of Neuro law topics And then finally we've got Jim Giordano who is a professor of neurology at Georgetown and chief of the Neuro ethics studies program in the Pellegrino Center And he co-directs the O'Neill Pellegrino program in brain science and global health law and policy He's also a member of the Neuro ethics legal and social issues advisory panel of the Defense Advanced Research Projects Agency or DARPA a topic about which I'll probably quiz him at dinner Which is a great opportunity And he's written literally hundreds of peer reviewed papers and several highly regarded books in the world of Neuro ethics So I think you'll agree. We've got a panel that couldn't be better to tackle the topic at hand so For those in the room, I've given we've printed up a handout for you that includes a glossary of Terms for the evening because we're going to be talking about Neuroscience and if you're from the law world, you may not be super familiar with the neuroscience terms And if we've got some of the legal terms that are involved if you're not from the legal world So hopefully that will help us. I've got a couple a second page that has a very very brief summary of a couple of Index cases in the law in the US. This is by no means a comprehensive list of actual legal cases in which Neurotechnology came to play a role and I'll just mention them briefly and quickly allow our speakers To begin so the first one is the the case of Dr. Lauren Sam Rao Who is a physician who was convicted on three counts of health care fraud? And he attempted to use Functional MRI evidence in an effort to demonstrate that he was being truthful when he said that he didn't commit fraud he had some fMRI tests performed on him and The director of the CFOs corporation that performed those tests Testified in his original court case and the judge ruled that the fMRI should not be used because it Lacked sufficiency scientific reliability and it might be prejudicial to the case The second case to just know about is the case of Terry Harrington who was convicted of murder He later after his first his initial conviction underwent an EEG based lie detection technique called brain fingerprinting That was being promoted by Dr. Lawrence far well And the the technique purports to demonstrate Electrically whether our subjects already familiar with a stimulus that's presented to him So, you know, here's a picture of the crime scene Did you know we can look at your brain and see if you recognize the crime scene that that sort of thing? And although Dr. Farwell gave testimony to the Supreme Court in the appeal The court didn't rule on the legal significance of the evidence because the case was subsequently dismissed on other grounds So there's just a couple of cases to keep in mind as we go forward and now I'm gonna let our first speaker come up Kate You are all set Hey, hello. Thank you very much for inviting me to this exciting event I'm really looking forward to the discussion that we're going to have afterwards and maybe even throughout So a lot of this work I did while I was a fellow at the Center for Law, Brain and Behavior over here at Mass General And I very much appreciate the support and I'm currently at University of Massachusetts Medical School So I'll touch on a number of topics and the first thing is really just to understand what my deception looked like in the brain Where do we expect to see it? We'll talk about three different meta-analyses that have been conducted on the topic Not gonna get into details But really just to see what are those patterns look like Finally, I'll talk about research limitations, which will be addressed repeatedly throughout this panel But I want to focus on two particular ones and the first one will be the populations that are being used and that is the absence of individuals with mental illness and the second limitation is the types of analyses that are generally conducted with fMRI research and then finally I'll conclude by summarizing What I've discussed and also hopefully offer some ideas for future research to people here and watching this So where do we think deception in the brain is going to be? Well, as you probably guessed, it's not one particular area but what you can also probably guess is that When we're lying we're usually Activating greater parts of our brain than when we're being honest and that has to do with cognitive effort And I'll mention that and this finding that there's greater activation in the brain During deception as compared to honest responding has been consistent in all 53 empirical studies that have been conducted using the fMRI and deception Why is there a greater activation? Well, it involves different types of Higher order reasoning you can imagine that you have to employ your executive functioning skills You have to switch between tasks to responding honestly and lying There's some moral decision component to it social cognition and really a number of other tasks All of which are happening at the same time and are adding to what we would consider Baseline which is honest responding But you can also probably guess that it's not one single area and I mentioned this at the beginning But really when we think about deception and more complex cognitive processes, we think of them as network based So we'll see if we can identify some of these networks that are out there So I mentioned that we'll talk about three different meta-analyses and meta-analyses is a combination of Different studies that have looked at the same topic where we analyzed the findings in order to come up with an overall conclusion The first one that came out in 2009 had 12 different studies and just to give you some dates the first study on deception using fMRI was in 2001, so it's a fairly recent research subject using fMRI's and there they use the activation likelihood estimate and it's really just a common meta-analytic Analysis form and they looked at a hundred and seventy-three areas in the brain that lit up more So during deception than honest responding then I'll point you to the Pharah and colleagues Meta-analysis that came out in 2014 it was published in Nature Neuroscience and that one as you could see had a lot more studies and finally I Recently came across this meta-analysis. I hadn't seen it before this is coming out of the Max Planck Institute It was published in 2014. They use a different type of analysis for Combining data and they looked at 30 contrasts so here's the first one this is the Christ in all data and It's not so important to sort of look at the different colors The different colors just symbolize the different types of studies that are included here But what I'm hoping you're noticing is that there's dots all over the brain, right? There's not this consistent pattern that one would want to see there's in the parietal region in the frontal Really, you could see some Subcortical activation Suggesting that it's it's sort of all over the place in some ways this was echoed by a more recent Analysis by Farron colleagues now this looks quite different than the first slide and that just has to do with How the researchers decided to present the data? But the similarity here again is that those activation points are all over the brain, right? We're seeing them in parietal frontal temporal And sort of showing this pattern or the lack of a pattern that we would hope to see if we were going to try to Narrow in and what deception looks like in the brain and This is the lassockian colleagues and again, they're presenting this data a little bit differently So it's hard to compare all of them just by sight But the idea here is that once again deception is all over the brain. It's not one particular region So the question then becomes what does mean, right? I mean we have activation all over the brain Can we try to explain why there's activation in the frontal the parietal? The subcortical areas and is there a way to combine that in order to say well Maybe this is what deception looks like in the brain and a couple of researchers have tried to do that So that Christ at all met analysis that I said that was published early on thought that one way to capture The construct was by saying well, what are the areas of activation for executive functioning networks? so things like working memory inhibitory control and task switching and I don't think this picture actually does a justice, but what it does suggest is there is overlap between deception and These other different types of networks that we often see but it's not great right I mean there's sort of a lot of areas that you see that are not being captured and Some of it has to do with the parietal regions and I'll address those in two slides But the idea is maybe some of this is being captured by executive functioning We also wanted to look at some of this data and we took the Farron colleagues data And we thought that maybe if we mapped it on these large-scale networks that have become more Important in recent years in explaining brain activity and particularly the control networks or the network that overrides General brain functioning the default network or the way that your brain operates When there's where it's not actually doing anything particularly and salience network that is picking up on important cues for you And we thought ah well, maybe we'll capture it with this maybe we'll capture deception And in some ways we did we captured a little bit of it, but again, it's not great, right? I mean just by looking at it you see it's not mapping on directly so Moving on from the networks. I thought Lesovsky and colleagues had a really interesting idea And they said well, you know what? Maybe it's not the networks Maybe it's the way that you're combining this different data And what you're not parsing out is the fact that some of these studies are looking at people provide Who are also engaged in social interactions when they're lying So they said well, what if we compared those lies that are in social interactions The ones that are probably much more important for the type of legal implications that we'll discuss as compared to non-social And what they found was in every single case There was activation in the brain in social areas the kinds of areas we would expect to see And there were no unique interaction activations in the brain in the absence of social interactions And Lesovsky and colleagues said that some of this helps to account for the missing pieces in Christ and colleagues It's difficult to tell because actual comparisons were not done But if you look closely at some of the regions you could potentially explain some of those differences So to summarize the meta analysis findings. Well The same thing that we knew before which is that there is greater activation during deception than honest responding We also noted that there are multiple Different regions and networks that can help to explain the cognitive processes that are involved in deception But the problem with that is that we still don't have a single answer about what a deception network looks like Partly because we keep seeing these very Dispersed areas of activation. We don't quite know what that means So in that light, we also want to look at some of the limitations, especially as we talk about applying these findings to Psycho legal issues and I mentioned the one that is particularly close to my heart as a clinical psychologist Is the fact that there's really a complete absence or almost a complete absence of individuals with mental illness And that is highly discordant with what we see in the criminal justice system The type of individuals that are actually going to be questioned So what we see is that out of the 53 different studies that are out there Only four of those studies have even tried to get participants who have some sort of a mental illness and by mental illness We can almost put that in quotes and i'll explain why So if we use the 53 and i've counted up the number of participants you'd have that over a thousand Only five percent of those had severe mental illness and those were individuals who had schizophrenia or schizoaffective disorder 0.001 It's really an embarrassing number, especially when we think about individuals in the criminal justice system Who have a disproportionately high number of traumatic brain injuries have had that And we had 56 participants with antisocial features and even that number is actually incorrect I'll tell you why so antisocial traits very important criminal justice system. I won't get into it. I think we can all figure out why and one of the bigger studies Conducted by fulham and colleagues That should say trade measured these traits antisocial with self-report And this is how they went to select their participants. They targeted university sports teams Participants because they felt that participants from these populations may show higher levels of subclinical Psychopathy so if you played any sort of a sport you would be Potentially one of the individuals who has high psychopathic traits Both of the studies the two studies have showed an inverse relationship between bold signal and antisocial and psychopathic traits This is critical because what it's telling us is potentially the people that we want to capture Are the exact same people that are going to show less activation and we will not be able to see them That finding finally my last point has to do with the types of analyses that we use Has to do with how we analyze fmri data We generally do very basic means comparison, but that's not what we usually do with medical tests What we want to do is we want to have classification accuracy Rates and only seven studies have done this. I won't get into the type of Techniques that they've used But when we do look at the seven studies and I'll just give you three examples of this We see that those classification accuracy rates are fairly poor Now I guess it sort of depends on what you think of as poor and that's probably up to the judge and the jury to determine And here you can just see there's a huge range of these and some of them range from 66 accuracy rates While others say that it's 86 percent accurate and I guess the question for all of us is What is accurate enough So to summarize The applicability of current fmri findings to real life settings requires much more work Especially in sample selection And the types of analyses that we're going to do in order to make sure that this is applicable to the legal system And in other settings And in my opinion, I think the future research should also examine not simply The classification rates, which as of now we're not doing But also to determine whether the existing technique the fmri actually adds something on top of what we're already doing And some of this has to do with psychological testing that is already out there to determine whether someone is feigning or not That's it. Thank you Okay. Oh, thanks. Do you want to stay there? I'll give you that All right, so next we have uh, dr. Giordano and I'll just have you come on up Sorry, I won't introduce you again. Although the first introduction. I had to cut out so much that I felt guilty Yeah, I think what he cut out was affected. I mean particularly looking at psychopathic traits that are subclinical. I'm a former wrestler So I'm beginning to worry Yeah, I know it makes me it makes me sweat So what I'll talk about here is the idea of Literally opening pandora's box as a pregnant pause as to whether or not neuro ethically We're at the point to use many of the exciting findings and techniques That kate elucidated both from her own work and from a variety of other studies most notably some of the meta analytics done by the max Plunk group and certainly Martha farra's long-standing work and whether or not we're prepared for that I don't think we're at a point where we can any longer simply regard a precautionary principle But rather one of preparedness Simply put I'll cut to the chase spoiler alert We're close. We're very very close And I think the predicate question that I'll raise is whether or not we have the techniques and technologies that are currently available To allow us to develop protocols if not a paradigm to reach into and extract from What may be a noise floor of various network activities in the brain to be able to parse from that viable signal That is in fact both valid and reliable for us to use in a variety of frameworks Including for example the law and increasingly if we look at other applications That have national security intelligence and defense potential there as well With regard to the title the pandora's pause I give great thanks to one of my graduate students mercedes sanat and also ongoing thanks to some folks who work in our group And this is tim brindley and kathy shots So I think first and foremost Let me post you some challenge statements as you heard earlier the question We ask with virtually all of these neuro technologies and techniques Particularly in light of the new brain research through advancing innovative neuro tech agenda Is how does the brain function and cognition emotion behavior of which certainly deception may be one? So the idea that the individual may have engaged in certain activities Through the truthful about or not and then how does the brain register these things? Can we assess these through the techniques and technologies at hand either alone or in combination? But more importantly if the answer is yes even to some extent How can these be operationalized the point that Kate essentially left off on in other words? What is necessary? And if it isn't satisfactory, does it at least? Satisfice I think a good term that it's sufficiently satisfactory So it's to be able to get the job done a point that I will then defer over to my colleague Dr. Professor shen so the issue there is what we do with the capability and Please take a look at the image that I provide for you here and here's your working caveat Let us not fall victim to something that we refer to as Icarus folly And this is the idea simply of false hubris What can we really do with the techniques and technologies and what are its limitations? And I think that's very important as the first initiative neuro ethical stance If we're going to say that we take neuroscience and look at this in a way that is ethical In other words, are we doing the techniques and technologies quote rightly? And then can we use them in the law? The purpose of the laws to instantiate some good within the polis Then are we also upholding the good as we transfer neuroscience into this particular vector? And that's certainly I think an important task if not challenge an opportunity for neuroethics So what we have to take a look at here is not only the techniques and technologies But the information they yield and here we get into the issue as you heard previously With many of the meta analytics many of the rank ordering studies many of these large scale Data depth dives that then extract information and try to do this in a way that is both reflexive and responsive Is it valid now the tenig's valid? Are they actually doing what they say they're supposed to do there are certainly some issues With fmri and other forms of neuro imaging and physiological measures Not least of which is that we're dealing with proxies We're not necessarily dealing with neurological activation per se But something that's second or third click off blood flow to brain areas that are activated with regard to particular Types of cognitions emotional responses Then can we take those and apply them reliably in various circumstances and situations So as to retain the generalizability Either taking group individual data in some way parsing them or taking individual to group data and parsing those And we'll talk about those momentarily and then ultimately what is the value of that information for the task at hand Does it translate well from the bench? Or the bedside into the boardroom of the courtroom and beyond and of course This is what we really get into the issue of contextualization of what we're doing with neuroscience A point that one of my colleagues john shook and I have made repeatedly in the literature One of the things that comes up with regard to this type of approach Is what kind of brain information we're looking at in virtually all the meta analytics you heard thus far What we're talking about a primarily group aggregate data that we can then utilize or leverage and gain some purchase to trying to understand What an individual's brain is going to do Well, i'm going to paraphrase my colleague hank greely at stanford and says that we don't bring groups onto the stand A point that of course professor shen will embellish upon at length but Could we Could we take a look at what many groups of brains that have been mean average across these types of studies have done and say Well, this represents a viable pattern and I suggest perhaps we may be able to do that The other is can we take a look at an individual's brain pattern? In other words the way we do lie detection and say that certain variations in those patterns and activity network Responses and profiles are in fact representative of groups of individuals who've done just this type of thing Either with regard to the cognitions that may advance deception or perhaps even their actions their behaviors That may be involved with things like culpability capability Capability of them to even inhibit their own behavior what we like to refer to as free will And I think one of the other issues there is that is a brain image taken today Representative of that particular brain at the time of the commission of whatever this act was Something we refer to as comparative decay And we have to take all of these things into some type of quasi or pseudo calculus and then develop from that A pattern or set of protocols that we feel will in fact hold the test of solidification And as a consequence of this be able to be used in these contexts Well, there are a couple of things we need to be very wary of first and foremost is the idea of picture thinking vittgenstein referred to this and the idea there is picture worth a thousand words even if the picture was wrong Again I refer to some of the work of nita farahani and grily Walter senate armstrong who've shown that in certain individuals what you do is you put up a really pretty picture of the brain Something like this for example and what we say is well that holds an awful lot of weight But I also caution you here along with false hubris. I think we also need to be aware of neuro nonsense I always wish I had coined that term But I didn't that's actually a term that the philosopher roger scrutiny uses rather fluidly and and I too as a neuroscientist and neuro ethicist Caution against neuro nonsense I think that there is some miscommunication and certainly some misinterpretation Of what we can do with neuroscience and its techniques and technologies But let's not be fooled There's also a lot of power in them their techniques Sometimes what we find however is this type of miscommunication is not necessarily coming from the scientific push side But rather is coming from a very strong pull side to be able to want very strongly to utilize this information in these results in particular ways And I think that there needs to be I think Directed in explicit stance from the neuroscience community from the neural law from the legal community To be able to formalize exactly those ways that neuroscientific information can be parameterized and used In some cases images such as the one you see above are rather valid in other cases. Please look at that carefully It's nuts Yes, folks, that's a walnut And as you can see, you know to the uninitiated people could go wow look at the clarity of that image Of course, it's a walnut And so what we need to really do is be able to not only sift the signal from the noise when looking at the brain But sift the signal from the noise when looking at the way we use neuroscientific information We have to be very wary of neuro nonsense of neuro babble of frank misuse but It's not throughout the baby with the bathwater Because there is great power in them their neuro techniques as I alluded to Are we there yet? Not quite But I would argue and I think I could argue defensively in the most Ignatian spirit What I would in fact assert others may deny but I'll defend this claim And I think my co-speakers on the panel tonight certainly are back me up on this We're pretty close We're not pretty close in ways that are miscommunicative. I think there have been some efforts at that I think we're very very close with regard to the factual reality of this But I think there are going to be some things that bring us ever closer not just in a tangential way But I think in a more precise way I think to do this there's a very strong neuro ethical imperative and let me speak that to you This is neuro technology being used both as a primary good. In other words, it is science and technology Which indeed has some primary public good. It is in fact in the public domain It upholds a public good as being able to advance the human condition and human endeavor And certainly this is one context of that and as an instrumental public good vis-a-vis the public good of the law We now super contextualize this and as such I think what we have to do is meet a couple of criteria first and foremost It has to be what we call authentic trustworthy to the facts Do with the neuroscience and neuro technology says it's going to do Delivers and can do work within the parameterizations of that and there's a lot we can do it has to be genuine In other words, it has to uphold the qualities and claims that we're making from this in such a way that are indeed wholly consistent Has to be efficient. It has to be effective in that way Non-wasteful we can't just waste these resources and use them in a way that's Willy-nilly just because we have the tools so coming to some technological imperative because it looks like to have pretty pictures But effectiveness is very very important in other words as you will hear it has to assist the trier of fact either at the point of the actual nature of guilt or non-guilt culpability and culpability or In sentencing a point that falls far outside my area of expertise and certainly within professor shens And ultimately it must also not harm because arguably the driving force behind the development and instantiation Articulation of any technology is for the good one might argue. Well, what about weaponization of same? But let's face it every country goes off to war thinking that a deity or some power is on their side We all create weapons ways of dealing with another because we want to defend what is ours And in this particular case the same is true When we're leveraging this within the law it must uphold the probity of the law and that's a context that is in fact solid Again that I will allude to my colleague professor shen to explicate in great detail But to do that it has to be what we call telecly aligned in other words the ends of the law and the ends of neuroscience must line up Which means the real test that falls incumbent upon the shoulders of people like me People like kate and others who do the neuroscience is not just to sort of align neuroscience with the law But to create directed and explicit discourse with those who do the law and say what can neuroscience deliver for you In other words not neuroscience in law, but neuroscience and technology what we called neuro s and t for law One way is to formally pattern What the necessary proxies are in other words if we understand that we're dealing with proxies that we're dealing by Reaching into a noise floor to parse signal these types of events these types of protocols must be formalized by consensus Both neuroscientifically and what is going to be valid within the scope of the law We require a large-scale data repositories and certainly the illusion to big data here is not lost These in fact must be curated. They must maintain some level of provenance But also they must also uphold the probity of what the law does so that we don't necessarily get into some problems with this That I'll allude to in a moment We need to embellish the science and technology in other words We need to fine tune the neuroscience and neurotech and we're certainly doing that through the use of something called a scientific convergent approach What I've illustrated for you here is how we may want to use individual data across one's lifespan To then engage cohort and group data in both comparative and normative way And certainly big approaches to big data are indeed oriented towards this and when I say big approaches to big data I mean it the idea here is utilizing multiple different approaches to integrate big data in a convergent approach that utilizes This type of thing The idea here is that utilizing brain imaging or other forms of neurophysiological and neurotechnological Orientations should not be a standalone but should be used within a larger form of multiple types of orientation Including but not limited to perhaps how we engage social behavior social dynamics and social media That may also gain some purchase or leverage within the law Utilizing a host of our existing techniques as Kate had said Fitting neuroimaging in as part of a larger picture What we sometimes refer to as a piece of the puzzle of that larger picture I think is important because in fact that is exactly the way we use science and technology But this too is not without issues and there she is pandora cracking the backs as suddenly We find demons that may come running out if in fact we know this is what's in the box How are we going to be prepared for the demons? What are they? Well first and foremost the issues of privacy versus protection We've discussed this in some detail in some of our recent work But I think this is in fact a balancing act Are we really ready to have the type of serial neuroimaging across various poisons a lifespan so as to be able to gain Baselines for individuals of what represents their norm individually Versus cohorts groups etc in various situations Are we really ready to take individuals back into the scanner as a consequence of what may be demanded by the law? We have to be cautious of bastardized use Anytime that we are going to use neuroscience in such a way as to make particular normative claims About individual states behaviors or condition This is a very preloinable enterprise And so I think that we really need to maintain the probity in all of its dimensions So there's not some misuse under a variety of factors And we have to be careful of something that we refer to as non harmless naturalism here with the deep nod of homage To robert almaider who talks about the fact that we can only use the facts that we have at hand, but We have to be very very cautious about the currency of those facts Are we using old neuroscience to uphold new ideas? Or will we say in this particular case that Fixity of purpose is going to not only demand But mandate flexibility of method and we will be equally flexible to then change particular admission standards if not the way we use things in law based upon new neuroscientific information So i'll leave you with a couple of quotes And I think these quotes are important to frame the work as we bridge from neuroscientific capability its gaps Its limitations and its delimitations Through the lens of neuroethics to be able to uphold some rightness and goodness so as to be able to essentially operationalize the power of these techniques and technologies in the context of law Come an Aristotelian framework We can say that every technique every tool and every inquiry And every praxis the idea of doing something in its pursuits is believed to aim at some good and this Incidentally no different despite what certain naysayers may leverage against neuroscientific and neurotechnological information being used in this Ways the idea is to try to uphold some good both privately as well as instrumentally And rand has come back and said that every horror of history was committed in the name of some altruistic motive that sends I think an interesting perspective on our first statement And of course if we are going to use this we need to determine how we use it And this must be a consensus that is in fact heavily weighed And laden with ethical probity as well as the way we're going to use this in a way that is not harmless to uphold The good of the law as it is defined and in this case I allude to the work of alistair McIntyre the philosopher at university of notre dame who asked what good What rationality and their organizations that are certainly posturing to this Much of the work of the mccarthur foundation International neuroethics society the international neuroethics network and certainly each and all of the groups that are involved here inclusive of the group of Harvard ours and each one of the ones that are represented by our panel here today So I think we are able to chart or at very very least view a path forward And to do that what I urge is the idea of utilizing neuro technology That's on the lips of everyone from the oval office all the way down to the smallest laboratory in neuro In a new way We're going to go forward. Let's go back to the future. Let's take a look at this term technology as neuro techni logos Let's bring forth that logos term a rational accounting of the tools The tool use and those of us who develop and use those tools and the realities of the human endeavor at hand In this case the law This is both a challenge and an opportunity With every act of increasing capability and knowledge comes great power And I think there's great power to be had here But as you know with every act of great power comes tremendous responsibility And this is both the task and the challenge of both neuroscience and neuroethics To opportunize the factual good and right use of these neuro technologies and ways that uphold the good of the law You want to get in touch with me? This is where I live and I thank you very very much for your time Jim thank you for admitting that that was a walnut because my differential diagnosis didn't seem very plausible What was wrong with that brain? Professor shen Why don't you come on up and tell us what to think about all this when it comes to the law and justice Well, I'll tell you a little bit about what to think about this Let me say a few thank yous first and get a couple things out of the way First I want to thank members of my lab adam steiner and morgan carlson and jordan christ for their work On this project. I want to thank above and the center and toss for making this possible And thank you the petri flume center and to clbb as well. It's been a wonderful semester here And this is great to participate in this program I'm going to try and do just two things today The first is allow you to walk out of here knowing where the polygraph stands and where brain-based lie detection stands In terms of its admissibility in the court and there's a real simple answer It ain't getting in for the most part. There's a more complicated answer and you'll walk out of here with that as well Then I want to think about the future of brain-based lie detection and whether or not it has a future at all In the courts and beyond and I want to suggest to you that the answer is possibly yes For some of the reasons that have been brought up before this is going to be an admittedly cursory overview So I want to get a couple things out of the way. I hope you get interested in this I hope you get interested in this topic and if so you can pick up This book which is ours, but there are 800 pages in it But 80 of them are devoted to brain-based lie detection and if you're really interested and this book is expensive So I can't advocate that you buy it, but I can't advocate that you come up after and get the free Website where you can get links to all these cases and articles by scholars that you've heard I want to mention two other resources one is lawnuro.org It's the mccarter foundation research network on law and neuroscience There's a searchable bibliography you can get lots and lots of stuff including great articles like those by a law professor William Woodruff just giving you everything you need to know on law And brain-based lie detection and then finally this is our lab in the summer We'll have an active website, but not quite quite yet our motto every story is a brain story But whether or not brain science ought to enter the courtroom of course is a contested topic So simple story polygraph admissible or not? No with these notable exceptions the state of new mexico Which treats things a little bit differently and in some states under some conditions for instance if both sides stipulate Prosecution says we'd like the polygraph defense says we'd like the polygraph or in a civil case both parties Plaintiff and defense say we want to do this and the judge has the opportunity in some states to say admissible But for the most part not and for the most part not admissible since 1923 That's when a case called fry was decided in washington dc It spanned its decision all of a page and a half So it did a lot of work in a few pages and it continues to be cited today And there are two reasons generally that courts tell you why they're not going to admit the polygraph Here they are i'm going to give you the two rationales. This is quoting from a Not so recent case, but i could give you hundreds of cases that essentially say the same thing And the first rationale is a science rationale essentially the science isn't good enough. Let me give you a representative quote These courts reason that the polygraph does not command scientific acceptability And that it is not generally believed to be sufficiently reliable and ascertaining truth and deception to justify its utilization in the trial process And in doing so courts are doing what they always have to do with any type of evidence Which is to say do we put it in front of a jury? We've got an adversarial system with lay jurors And so we have this process that you know, for instance many courts on the continent don't when they have Judge-led and judge-centered investigations. So the first reason straightforwardly enough and we'll come back to this is the science Just isn't good enough But there's another reason unique to the polygraph that you won't see come up in for instance Types of some of the other expert evidence and this is the fear of over persuasion A famous quote that's sometimes cited is that the jury is the lie detector And here i'll give you again just a representative quote from a case Quote aside from doubts about the reliability of polygraphy at this stage of the art There are some countervailing policy considerations A foremost concern is the effect that such evidence will have on juries provided that the polygraph attains The degree of scientific acceptability and reliability Which will allow its admission in evidence and here's a quote I particularly like quote when polygraph evidence is offered in evidence at trial It is likely to be shrouded with an aura of near infallibility akin to the ancient oracle of delphi You know you imagine jurors and here comes the expert with her or his polygraph Or now his or her brain-based lie detection evidence and the jurors faint and they say okay We don't need to hear any other evidence because we've had it all there I'll come back to that line of reasoning. I think it's actually empirically Consentious, but that's what you'll get for most polygraph evidence So what do you get with most brain-based lie detection evidence? Well, we only have a few cases, but the answer is again not admissible And for anyone who's been following these things that outcome is not at all surprising I had an opportunity yesterday to sit down with steven lakin who is the expert who testified In a case called usv semeral. Let me just give you the 60 seconds capsule of the case You've got it there, but in case you haven't seen it loren semeral runs medicare and medicaid company In uh the south and it's pretty clear from the billing records Which the government prosecutors obtain that he's over billing you submit this code when you should submit that code and oh You get a little extra money He's accused of fraud and in order to be successful The government must prove beyond a reasonable doubt that he did this fraud knowingly Part of his defense was well It's really complicated to do these codes and we called the 800 number and yet you got me dead to rights That we over bill but I didn't do it knowingly And what was novel about this case is that his attorney houston gordon sought out Dr lakin and they wanted to introduce fmri based lie detection evidence That lock dr lakin would get on the stand and testify That in fact dr semeral's brain was telling the truth when he said that he didn't know that the fraud was going on And so a dowbert hearing was held a dowbert hearing is the phrase we use in the federal courts for assessing Whether or not expert evidence should go in front of the jury and you've got there the factors that are considered The crib sheet essentially is it good enough science has it been reproduced peer reviewed known error rates and the like And the judge decided in that case and this is the quote At least in this early stage in this development effort mri based lie detection does not satisfy the requirements of 702 We've seen some other cases like that and it's a good bet that it's not going to be I imagine admit it anytime soon if the polygraph hasn't neither will brain based lie detection at least as it's Currently presented so there you go That's the simple story about polygraph lie detection and in fact at this point you might be wondering Why have this discussion then if the polygraph hasn't gotten in for the most part for a long time And if already we've had some cases and brain based lie detection isn't going into the courts Isn't that the end of the story seems like law is kind of standing up With its walls and not letting this in so let me give you the complicated part of the story Why do you think it is that if the polygraph is generally not admissible in courts? We have over a million and some estimate 2 million polygraphs administered annually in this country Where are they going? Well ironically enough a large percentage of those now some are in the private sector But a large percentage of those are being carried out by the government And a large percentage of those are being carried out in the criminal justice system where in police investigations And so this very technology that is deemed as you've seen not worthy enough for courtroom used is used regularly In all the stages leading up at least some of the stages leading up to the courtroom And he's used regularly in government employee screening and re-screening in a number of areas and there have been Actually debates between the national academies of sciences and department of energy and others about the validity of using And for the fbi and cia use of these technologies But there they are And so let's pause there because if you said well, I wish that brain based lie detection would just go the way of the Polygraph let's keep it out of courts Are you also then going to suggest that there should be a million brain-based tests a year and that it should Be become part of the arsenal of police investigation And that it should become part of the arsenal of department of energy screening and the like And I think the answer to all of this which picks up on what our two speakers have already said is It's an open question It's an open question in part because we don't know if there's value at it or not We don't have that research yet But again query whether we might get there someday Let me move to the future now and in the time remaining I want to just make some points and raise some questions So point number one is that dahlbert and rule 702 and the federal rules of evidence Aren't the end of the story. I'm going to dwell on those in a few minutes But there are many ways In which you might use this evidence even in a courtroom and not have to confront Here a couple one is at sentencing in the united states. We bifurcate our criminal trials We don't have to do this some countries don't but we do the jury or it was a judge On the it was the fact finder makes the determination of guilt or innocence That's why in the movies they come back and say we find the defendant guilty They don't then go on to say and we've decided to sentence him for 15 years No, they don't do that the judge does that and when you switch to the sentencing phase accepting capital cases in a few states There is some jury sentencing. It's then the judge who decides what the punishment will be And when the judge decides what the punishment will be she or he can use a vast amount of evidence that wasn't admissible Even at the courtroom stage and so for instance might a judge want to know and might neuroscientist be able to give us some evidence That tells us whether or not this individual at this current state of time not 10 years ago feels remorse Or is she or he lying about whether or not he feels remorse might that be useful to the judge might that be something That neuroscience could offer who knows but the point is that there are avenues Even if it doesn't reach the dowbert standard Another would be again at the police investigation stage and of course in these many non courtroom Uses second point to make is that I think there's actually an interesting amount of growing empirical evidence Some of which we'll publish soon and some of which is in are in the literature already That this concern about the oracle of delphi is just not warranted jurors I think are smarter than that especially when you think about what dowbert itself recognizes are the two primary ways We deal with shaky evidence in the law. They are not to keep it out of front of the jury They are to allow for cross examination If you are an expert or have ever been an expert You know that cross examination can be a very difficult thing to go through And the calling of other expert witnesses And so the question is not will you faint when you see a brain based lie detection evidence present? But will jurors decide any differently when that is presented when there's cross examination when there's another expert Who says aha, that's all garbage and I think the answer is likely They are not going to faint is which isn't to say brain based lie detection might not have an effect But he has to say that that particular concern may be a little a little broad I do think there are some real concerns with the science written on this we've had this discussion already will have some more But I want to clarify and this is sort of the big final point to make that what the law requires And this is what kate asked how poor is poor and how good does it have to be What the law requires in many cases is very little you have a handout It doesn't it has a rule 403 I'm going to read you the other two rules the lynchpin of admissibility for any type of evidence Mr. A has a case and mr. B wants to testify the lynchpin of is relevance And here's all that it means in law to be relevant evidence is relevant If it has any tendency to make a fact more or less probable than it would be without the evidence And the fact is of consequence in determining the action we call that probativeness and materiality and rule 402 tells us that if it's relevant It's admissible unless there's something else that gets in the way This is why really shaky witnesses who everyone in the room thinks they're a horrible witness We think they're lying through their teeth. This is why the judge doesn't say okay, mr. Mrs. Witness, you don't get to testify You can go on the stand through cross examination and through these other witnesses the jury might well say Witness be boy. We think that she or he was lying So there is uh that opportunity sort of that thrust of the federal rules is to put more in front of the jury There's a special exception however for expert evidence rule 702 And dober which does require a higher bar for I says it nicely They have a phrase the twilight zone of evidentiary force and you're never quite sure when it's good enough But let me give you some quotes from the semaphore case so that it's clear what dr. Lakin was actually Trying to say and what he was not he was not contrary to some Even in the academic literature that I have the dispositive answer about whether or not sam Rao was lying when he Lying about this fraud in fact He says quote It's not up to me to make any sort of statement or any sort of opinion about whether they are lying or telling the truth You might say well, wait a second. That doesn't make sense Well, here's how he explained in bayesian terms and I think of this in bayesian terms as well He says this guy and this is what he says on the stand this guy comes into my office I don't know him from adam 50 50 chance. He's lying. I put him through my procedures with I trust I then interpret that data and he used this percentage well I think now 57 chance. He's telling the truth Would lakin said at the end there was a question Well, do you think that jury should rely just on this? He said no Here's his quote if one were to use this technology i.e. his f from rye technology As the sole basis to make a conclusion either positive or negative Then that is an inappropriate use of the technology It needs to be a compilation of forensic evidence The jury can decide how important those things are in determining dr. Sem Rao's guilt our innocence That's a very different approach than again if you just hear oh, it's a brain based lie detection case She or he must be going in saying I've got the ultimate answer It's a very different approach now whether or not and some might argue you can't even say that You can't go from 50 to any percent because it's so inconclusive that it doesn't assist the trier In fact, that's an open scientific question But I think this is kind of where we're at right now We know that for forensic science community doesn't have to be perfect indeed in 2009 national academy report Really was critical of a lot of forensic science bite marks any fingerprints any number of things Those things in fact it was there was great hope at that time that there might be a re examination of those types of evidence And they're probably ought to be but right now those things are routinely admitted despite that the fact that they probably don't match some of these dowbert Criteria, but it's thought to be they're good enough or at least we've been doing it for a long time So you've got that in the background And here we are we know it's not dispositive. We know it's at least not junk science What do we do and does it begin to look a lot like other evidence if it looks a lot like other evidence? It's not so special, but maybe it's also not so scary for courtrooms Well, thank you to all of our speakers We've got plenty of time for Questions and answers. I'm going to bring the mic around partly so that the people who are watching via webcast can hear your responses The mic is is going directly to that cast And matt bowman over here is monitoring the twitter feed matt you can you say what the What the twitter handle is to send in a question or is that obvious to anybody who's At hms bioethics so if you're in the on-institutia to speak up you could even tweet matt a question and we'll we'll try and take it Is there anybody who has a question right off the bat? well, I'll Let me let me take this question The human brain is pretty good at deceiving other human brains Do you think that when you develop your? Functional mri or whatever technique you would you use in court That the human brain can learn how to deceive the technique you develop Are you saying? Are there ways that the human brain or the individual can think to fool the fmri is that? Yes, and the answer is yes There are many different ways And part of the problem with fmri is in particular in your imaging Is that there has to be a lot of precision in the way that person? Lays down the way That they are responding and part of that problem is that some individuals particularly those with mental illness Those that are more likely to be in the criminal justice system Are going to have a lot more head movement So in some ways it's going to be even harder to accurately detect now that could be purposefully or that could be just Naturally, but we also know that there have been a few studies not many Think too, but I want to make sure that I sort of give you so a few studies that have looked at Counter techniques that you can take in order to make sure that you are actually not Experiencing what you would expect to experience when someone is deceiving so you could sing a song or you can think about different things So yes, those things are out there I think we have so many other problems I'm not sure that that sort of The primary issue right now, but it's absolutely one that we need to consider especially once we started talking about sort of things like defense and sort of going into Defense work and terrorism, etc So I have I have one that kind of follows on that question a little bit So if I'm thinking about all the neuroscientific problems with fmri based lie detection techniques, I can I'd probably have to use both hands and a foot to count them up But the one that makes me think that that this might never be possible in practice Is the is the difficulty in defining deception and then figuring out what the neural signature of that is right? So if I'm a graduate student and I'm asked to come into the lab and the researcher says I'm going to show you a card. I want you to lie about about the card to me Well, am I am I telling a lie if they asked me to tell a lie and also there's no stakes involved Right, so that might be very different than what happens if I if I really have to deceive the person I'm talking to And so I'm wondering I'm wondering if you could say anything Kate about that or Jim About about the the neuroscientific difficulty in in actually detecting lying Let me start the conversation and certainly hand it back over to Kate because I think it needs to go down the rabbit hole a bit That very point was something that had been brought up about Five or six years ago. Actually when there was a military effort, particularly with regard to intelligence To examine the possibility of utilizing brain-based deception detection techniques And although it was not necessarily abandoned It was thought that those techniques are just far too nascent at that time and were not able to really parse out True signal from what may be noise Particularly in light of your questions or as to whether or not this might be able to be faked good or faked bad Given the existing neuro technology that was available in 2009 So at that time it was thought that other techniques might be viable And certainly the notogram that I put up with regard to utilizing a whole host of different scientific approaches As professor shen alluded to and that actually came out of some of the work in the semeral case Seemed to be the default in other words. We need to put all of these things in a greater context But I think a larger issue there is also Whether or not the technology can be combined in such a way So as to sidestep if you will the deception detection bullet in other words We're really not using it for deception detection But we're using it to evidence other forms of cognition that may play a role in individuals Moral intuitions and perhaps what we then view to be moral ethical or even legal Capability to act or not act within the scope of whatever we define to be the law I think there it gets a little more into a gray zone because I think some of the work that has looked at certain Studies of first or third person processing of what we view to be Moral or ethical decision-making or ethical problem solving Has lent some particular credulence To the capacity of these types of technologies to at least identify networks involved in this signal And that's an important first step So I think to speak to your first question explicitly the answer is that That there are so many moving parts there that it may be a sticky wicket I think coming from that however is the fact that the neuroscientific community and other user communities Have been able to see that as a potential gap Identify the gap assess the gap and then look to delimit or compensate that in some ways While looking at other potential brain mechanisms involved in cognitive processing that may be equally as instrumental Okay so I completely agree. I think just one point that I do want to add to this is this idea of combining multiple methods and In theory or in some ways that's a great idea, right? If we can get a number of different techniques The problem is the reliability of each of this Technique is it is questionable. We don't have great numbers. Just like I said, we have all these Ideas about what deception looks like versus what honesty looks like We don't have a reliability for the bold signal to differentiate those two and when we do there's variable numbers So now you combine that with even say the basics of the polygraph Which is all over the place ranging from like 50 plus to some people report in the high 90s And mathematically that's a problem, right? You're not actually getting closer to the truth But in some ways you're getting further away because you can't add two things that are not great and get something good And that sort of a stalemate that I'm not sure we're going to be able to move forward until at least in some ways We get a little bit closer to getting those classification accuracy rates Much higher than where they are if I could just put a bow on that because I think that that's perfectly said I alluded to the idea of utilizing a convergence approach and I illustrated that actually one of the proposed Methodologies is to engage a process which has been formally called advanced integrative scientific convergence And the ideal side of that is that what that then does is desilos many of these techniques and allows for an opportunistic Leveraging of all of their positive points It's a two-sided coin the other side of that coin is what you then also do is increase the problem space So we can play tennis with this all day and say well, then you have a larger opportunity space to address and solve the problems Well, that's correct But even in the notogram one of the things we noted is the probabilistic inference of utilizing these different techniques together The more techniques you get the larger the probabilistic inference for both error as well as certainty And so those things have to be at least accounted for if not in some way compensated for and that is very difficult to do Basically the confidence interval of anything is going to increase to such a degree that we're going to start You know, we're going to be from 10 to 90 percent certain and that's not particularly helpful Thank you for Really thoughtful and stimulating discussion. I'm sorry that I missed probably the whole first half And so what I'm about to ask may be spurious but the part of the issue There are a few different issues and we've had a previous seminar looking at the issue of culpability and brain lesions and what you were talking about almost goes down that rabbit hole another sort of intermediate issue is the one of psychopathy and the changes in the brain and orbit a frontal and related systems processing That have to do with the The mental states And the behavioral states that allow certain people to even do things that would be unthinkable for other people in the first place And they're hardwired differently if you will in terms of their salience networks in terms of emotional reactivity In terms of arousal to certain internal and external Representational states etc and and consequences. So has there been What I missed and I'm now I don't want to ask you to repeat it, but I can go and look it up is what's the state of the literature about I know that literature at least some of it, but of the the lie detection with fmri and and are people taking those things into account And stratifying based upon some of these subgroups within the population So I spoke a bit about that and I'm more than happy to repeat it partly because I think it's a very important point And the point that I was making is that the existing literature that we have on Lie detection research, which is which is not small at this point almost omits All of the types of populations that are going to be found in the criminal justice system And as I mentioned, there's only two studies that I found out of the 53 And you know if anyone knows of any additional ones, please feel free to let me know These studies are not even necessarily Evaluating the types of people that truly would be in the criminal justice system. So they have anti social traits, but as Jim has pointed out Everyone potentially has those and the example By telling us but but The studies did not capture the concept I think the way that most researchers tend to capture the concept of anti social traits and psychopathy and even sort of in this Fairly ambiguous way that they've captured anti social traits What they saw was that there's an inverse relationship between the bold signal And that's what we see in other studies of individuals Anti social partly because they're not responding in similar ways And there if we were to ever get that deception signature as if you alluded if we could ever get there Are we even going to be able to use it for the types of individuals that we would potentially want to use it with? Let me build on that and then actually hand this off to to francis because I think it really falls squarely in his sandbox Is a new study that will be Released at the upcoming neuroethics meeting in october And it's the lead author on the study is Sarita silvera And it was done in the laboratory of christina henning fast at ludic maximans university in munich And they looked at individuals who were diagnosed along the anti social behavior spectrum Who presented positively on the psychopathy checklist versus normals and their ability to perceive guilt And how to they process guilt both in the first person as well as in others And also with regard to what they then viewed to be normal and there were differences and they were reproducible differences Now obviously a fairly small-scale study. This is being leveraged as an abstract But it builds upon the existing data that kate alluded to earlier and then Obviously those those data were presented to you if you wanted to read them in more detail But I think where that really plays a role is not necessarily in whether or not the individual is culpable But speaks very strongly to francis's point about what that then does for sentencing This gets back to the the panel We had a couple of weeks ago that looked at the whole idea of my brain made me do it so to speak Which no one really regards as as valid But certainly the possibility of whether or not there are mechanisms that mediate or involved in what some have referred to as Free won't and then what that means in certain conditions not necessarily disorders Particularly when one takes a look at the anti social behavior is not a frank condition Yes, no is a binary but existing as we heard earlier along a potential spectrum Which we both got a chuckle out of because collegiate athletes seem to rank on that and that was a collegiate wrestler So that worries me francis I would just add that I think a question for admissibility is whether it's bad enough That we don't have this evidence in front of the jury at all Or if you allow it in like we allow in a lot of other really bad evidence expert and non expert With the assumption that's usually played out that the first thing the expert the the lawyer on the other side is going to ask is And and how many of these subjects were psychopaths like the defending right and then let the jury play out and that's I think it's a challenging question even once we acknowledge what the deficiencies are whether they're sufficiently deficient To keep it away from the jury entirely Or to put it in front of them and lay bare via cross examination in another act was great question Well, I guess I'll go next Thank you. It was a fascinating discussion Most of this has centered around the limitations of this technology. So we've gone through a number of the problems Tosses Fingers and toes and and I'm sure more than that But jim you you said that Even though we're not there yet. We may be close And so here's my question If at some point in the future Somebody develops the perfect lie detection machine the perfect machine Um What is what does that machine look like? What can it do? What is the kind of evidence? That you would want to see where you would go This is it. We have found it. Um, this this does the perfect job um So instead of limitations what one of the positive things that this machine has to have that will convince you It's really working So since you asked me the question, I'll I'll I'll start it But then certainly I think it needs to go to kate with regard to what she would want to see coming out of the study She's doing with with fmri And then absolutely it needs to go to francis to say is this good enough Which is essentially the question that he leveraged and I think very nicely created his platform From our side, um, which is balancing the neuroscience and the neuro ethical component Which is I think what you're asking is it would need to certainly be valid In in all of its ways it would need to be valid as Uh, a particular construct workable if this is the way the neuro technology works The nature of the proxy would need to be close enough and be accepted to suggest that even though there is this proxy If we're using bold or if we're using some other signal the signal is wholly representative and there's not a noise Differentiation, so there's not a large gap between what we hold to be valid signal reproducibly And some other underlying event that may downstream evoke this signal that there may be Artifactual number one number two it would have to be as as I mentioned in my talk reliable In other words, it has to be generalizable to the specific contexts of use and do so in a way that is reproducible Time and time again given particular parameters that would need to be defined and defining those parameters. I think is important But then also it has to be shown to be valuable for that So it's not just a question of having the scientific tool But that the tool actually holds value I could take a look at a rock And a rock would be a wonderful tool. Yes, I could bang nails with it But it may not necessarily be as precise as I need it to be particularly if we've criticized other versions of similar types of things a plank A slate of doing the same type of thing at the same level of fidelity in the same level of Satisfactoriness or non satisfactoriness So I think we would have to demonstrate is that it sharpens the tip of the proverbial spear in some ways The three parameters that I would want to see is the validity the reliability and the value of that and in each one of those Parametized dimensions that are agreed upon through consensus Not only a neuroscientific community, but also the communities of use Is it actually doing what it delivers? Is it authentic? Is it genuine? And this is why what we're calling for is neuroscience for law Rather than just saying we have these techniques. Can we perhaps fit them? I think that the demands would be fairly high and not just for law by the way I think we see that in a variety of other things the same type of discussion for example could be Could be fostered over things like the use of transcranial electrical stimulation a variety of the interventional techniques Because I think even a discussion such as this Opens pandora's box a little further that says if in fact we're looking at brain anomalies or patterns of brain function And what we're looking to do is utilize them in the law Does this then at least provide a foot in the stirrup to say that we then may be able to utilize some form of Interventional neuroscience and technology to correct the rehabilitate and of course, that's a completely other question I just thought I would curl people's hair with that kate Well, I don't have too much else to add in terms of building a perfect machine um I did sort of start to think a little bit about so the issue of reliability and validity is of course Crucial to any test, especially as we start to think about it in the legal realm but the notion of Having tests that don't start out perfectly and either get better or phased out is fairly common So the idea that comes to mind or thought that comes to my mind is the Rorschach Right. Um, there was a point when this was a test that was used All the time now, we know there are a ton of problems with the Rorschach, but in most part it's not frequently used But interestingly enough it is still at times admitted into court now not often, but it is So in some ways there's been an abandonment of the Rorschach It's not part of the standard repertoire for doing evaluations But on the other hand, we've come up with new tests that potentially Try to get it the same ideas that the Rorschach was getting it So i'm not quite sure what's going to happen with this I think it'll be sort of interesting to see which route it will take Bob ox, I'll give you some sci-fi stuff that addresses exactly that question. What what would that technology actually look like? um So I wanted to ask the question about something that I didn't really see brought up tonight, which is the very high cost of brain imaging And so I wonder if we ever get to a point Like professor shen mentioned where we use brain imaging based lie detection As kind of normal evidence and we treat it in the same way What kind of Methods can could we have to kind of prevent it from being used? um It seems like it would only because of the high cost it might only be used in cases where a defendant wants to prove themselves innocent and prove that they aren't lying about something whereas So somebody else who is accused of something who may not have those opportunities May not be able to purchase a brain based lie detection for themselves How will we kind of Ensure that it is used in an ethical manner in a way that doesn't Kind of Already exacerbate existing socioeconomic problems within the criminal justice system. That's a great question It's been called a rich man's defense not just this but actually the stuff that was talked about before and It's not a coincidence that lauren semrow The setup was that he ran this large company. He had the wherewithal to do it and and the attorney to do it as well By way of numbers, uh, you can that type of thing cost under 5 000 right now So it's not prohibitively high for those who are already paying say $50,000 or more It's certainly prohibitively high for your standard criminal defendant. Um, who has hardly enough money to do anything So I think it would have to be State funded somehow which in the current system is not very likely to happen I will say and I've been waiting for the right moment to say this I think this might be it that um eeg based approaches, which are slightly different but exceedingly different on the cost side Might have some value here that we didn't talk about the far well case But it was a case where eeg based p 300 wave memory detection was used the logic and I actually think there's More future there is that under the right conditions where you don't have contamination That's a really big issue But under the right conditions, there's a pretty reliable very reliable finding in the lab at least that you can identify based on Eeg data whether or not individual is seeing something for the first time or whether they've seen it before So you train up the lab on the words is it right so the idea in that case was there was something from the crime scene that That if you were at the crime scene, you would have seen it would have registered There are a lot of chain of inference steps which are problematic in any case including this one But the basic idea is that that might be used in some context And if that approach were used to be a lot cheaper, but uh the the brain based fmri approach It's really expensive. It's going to be expensive for a while And as we've talked about here, it matters who you call over the expert who analyzes the data And and everyone can't afford the people in this in this room for sure, but great question I don't think we have a great answer yet so The way that the fmri and not specifically to lie detection but sort of in the criminal Justice system in general has often been brought up is To show whether someone it has psychopathic traits or mental illness And I know that there are a few appeals where the judge has said oh because you haven't used fmri. You haven't used neuroimaging They've granted appeals. Is that something that is now being funded more commonly by the state or or how do those tend to work? So It's a great question the term of art in our world is ineffective assistance of console So the defendant says I want an appeal because my attorney didn't do the things I asked them to do Now defendants often do this and in fact the attorney did do everything she or he could do But one of the things they often don't do that now some defendants more than before want done is brain imaging There are a couple of there are a few cases where the judge has said essentially You should have had the imaging done these cases are the exception to prove the rule And if you read some of these cases you see why the imaging should have been done Massive brain trauma multiple counts of brain injury for your standard run-of-the-mill defense You're not going to get brain imaging frankly. You don't need it But and the attorney doesn't have to worry about Probably that they should have had it now if that were to fall If the price were to fall if you know mgh opened up a shop and said come on in for $150 Well, then we're maybe in a different a different ballpark, but at least right now There's a little bit of that but not not a lot So this goes back to the you know the telos of law and the telos of neuroscience and whether those are aligned And a lot a lot of the panelists you've brought up you know research That you know questions that need to be asked need to be answered and my question is you know Since we don't have a branch of applied neuroscience to the law in the same way that we do to Let's say medicine. Who's going to do that research? And how can we you know create institutions to get that sort of applied research done? Let me start that thread because I think All of my fellow panelists can address that I think we're on that road I mean, I don't see that as something that's completely far into the to the current trend I think we can look to social neuroscience. Certainly. I think we can look to cognitive neuroscience The point that that kate made that deception deception engages social consequences speaks very very strongly to hear is a cognitive mechanism That's already the focus of much of cognitive neuroscience that is then being engaged in various social situations So conjoining those two fields We look at neuroethics Something along the kin of of neuro ecology in other words What are neural mechanisms that function in various ecological interactions for the housekeeping of the individual in context of their group in various actions So I think the neuroethics community particularly with regard to the neuroscience of Protomoral cognition is also gaining some momentum and steam and very strongly Interest in moving in these directions And of course all the work that's been fostered in the field of neuro law Has I think thrown down the gauntlet and said what can be done? So I think that there's already moves afoot now It's just a question of catalyzing them in such a way as to make sure that these reactions Are done appropriately and with the necessary rapidity, but prudently I would just say I think I would agree with Jim and hope that we do more things like this So that we have more funding at the interdisciplinary level, which doesn't exist a lot And we are seeing some movements, but frankly none of my colleagues for the most part know what neuroscience and law is Law students and law trained people are not science people on average and not math people And I think that hopefully that changes generationally as well So I think we've got time for one more before we move on to dinner I think it's fair to say that we can probably accommodate anyone here at the dinner even if you didn't RSVP I think we probably have enough food to feed you all so you're all welcome to come and join us afterwards But we'll take one more before we Reconvene at dinner I've heard a little bit about potential privacy issues and fifth amendment issues regarding brain based techniques in the courtroom And what kind of comes into my mind is if you can tell if people can see an unfamiliar scene if they've seen it before Is that infringed upon your right to not self incriminate yourself? So it's kind of wondering Where you see this conversation going if we get good technology will people be able to compel others to take brain imaging? Just your thoughts on that A great question. There are fourth amendment search and seizure issues and fifth amendment A due process is It boils down essentially in the courts not quite to this question, but almost to this is it physical or testimonial evidence So some have described brain imaging and EEG as brain fingerprinting And you don't have to consent to your fingerprints You're still going to get your fingerprints taken the thought is that it's physical You're not testifying you're not communicating anything on the other hand It might also be fair to characterize the brain imaging or the EEG data as testimonial because in fact You are communicating something perhaps under under duress. So courts haven't confronted this yet. Thankfully, there are some scholars Thinking about that on the legal side mark blitz need of our honey others and I'm glad they are And it's the legal community. I think is still working out what court should do And I hope that when courts confront it they listen to those debates, but it's a great question We don't have an answer yet Let me add But not much to that You know, I was being the the contextual drum over and over again I think what we're really confronting here is something that we've referred to as context drift And let me give you an example There has been a very strong movement to suggest that if we're going to move into precision medicine utilizing neuroscience Neuroprecisional medicine or personalized medicine based upon the adage that c1 brain c1 brain And that brains do in fact develop somewhat differently structurally and functionally That one of the ways that these novel Neuro technologies can best be employed is to engage longitudinal data across one's lifespan at key opportunistic windows of neurological development That would then provide a baseline index across various points in the lifespan So this is lifespan point data and point reference data And on one of the graphs that I showed you it looked like a funnel and we saw lifespan data That can actually be considered various points of one one's lifespan But we want to take structural and functional images of the brain doing certain things This is also a foot because this has been at least posed by and in some cases engaged by the military for individuals Who are off on deployment So as to refine a baseline measure against the possibility of traumatic brain injury Even if that traumatic brain injury is mild particularly want to be able to reference certain functional indices So again the idea of whether or not we do this type of thing falls into a number of different contexts It need not necessarily be gained with the explicit purpose of utilizing this data repository for the law But if in fact one of the things we're looking to do is to move into big data That then allows both precision within the neuroscience to be able to Essentially exploit in the most positive ways Different types of information from different forms of neurological assessment, whether it be Neuro imaging and all of its different forms physiological measures neuro genetics, etc Then what you're going to really need is a large-scale database And this certainly has been proposed The probity of that database whether or not that database is then accessible under the scope of law I think is certainly up for discussion as francis directly mentioned And I also think that you may be seeing a changing tide as to what represents the privacy factors and protection factors And this can come down to a cake and eat a two phenomenon that I think Will have to be addressed at some point in the not too distant future And if I were going to predict how far out I'd say within 36 to 60 months All right. Well, I want to take one more chance to thank our speakers for an excellent presentation. Thank you