High Performance Medicine with Eric Topol MD -The Front Row at Scripps Research



thank you all for coming my name is Jamie Williamson I am the executive vice president here at Scripps pleased to welcome you here to the front row I see lots of familiar faces that are my colleagues and people I've met before but I'd like to maybe have a show of hands who's whose new first time at front row look at that okay terrific well welcome to Scripps maybe I could just say a few things about Scripps you know we are one of the largest not-for-profit biomedical research institutes in the country we have two campuses we have this beautiful campus in La Jolla we have a sister campus in Jupiter Florida here we have about a hundred and fifty faculty members that span all kinds of disciplines through chemistry and biology neuroscience immunology structural biology and molecular medicine and we have about 50 faculty members on the Florida campus one of the crown jewels of Scripps of course is our graduate program we have a really fantastic set of 200 graduate students here and about 50 graduate students in Florida our program has been ranked in the top 10 in the country for about 20 years now by US News and World Report so we compete very effectively for top students all the way around the world we have in addition to basic research we have a few Institute's within Scripps one of them is the caliber Institute so Pete Schultz our president founded caliber about five or six years ago with a substantial gift from Merck and when he became the president three years ago he subsequently merged caliber right into Scripps so we now have the ability to do basic research and then translate that directly into medicines and the the licensing revenues from that will come back and fund some more research so the other Institute that we is the scripps research translational Institute and our featured speaker today dr. Eric Topol has was founded that Institute and has been running it and it's it's it's a another crown jewel at Scripps Eric is the master of genomics and high-tech and artificial intelligence as it is applied to medicine he was awarded the largest NIH grant that was ever in the wordid in the history of the NIH which is the all of us project that is in enrolling a million patients in trying to collect their data and trying to improve the way that healthcare is done and I'm sure he's going to say much more about that so I won't I won't say anything more but I will say that Eric is clearly a thought leader in the field he just has this way of distilling down information and then projecting it into the future what are we going to do with all this genomic information how are we going to use it and framing the discussion and the dialogue about how we need to change healthcare in in the United States at this time I'd like to welcome Eric to the stage and please join me in in welcome all right well thanks so much Jimmy and thanks for coming this evening for the front row especially the people in the front row here yeah we're gonna have fun we're gonna get into a new form of medicine it's a little bit of back to the future some of you will remember 30 40 years ago when medicine was very different when there was a precious relationship between doctors and their patients and perhaps is now a new mechanism we can get there and they wouldn't that be exciting so that's deep medicine in a nutshell but we're going to talk about how we're going to get there this high performance which is a take-off from high-performance computing of a new way to take care of all the data that's flowing now in every which way to help patients so this is a cover cover of a couple of recent magazines journals one is about being human an amount of data that we can see we can work within any person now which is extraordinary and then the whole idea of a data-driven medicine which hasn't been the case really and just starting now in many respects so it turns out that we've exceeded the number of bytes yottabytes and now there's a contest who should be the next unit the next one beyond yottabytes some have recommended hell of a bytes but this is just to give you a sense of how much data is being accumulated and by the way out of these were a remarkable number of exabytes in healthcare a lot of that is from images and also from sequencing so the first thing is to fess up about how medicine is really very shaky today in the u.s. we have 12 million minimum serious diagnostic errors a year and that's a very big number and the reason for that is because there's not enough time to review data there's not enough cognitive aspects because things are so rushed and there's a book that many of you have read the any condiments really now a classic book Thinking Fast and Slow but we have a different form in medicine today medicine fast and shallow and it's very shallow not just about time but we just can't as doctors and nurses clinicians can't get our arms around all the data about any given person and so that has to change this is a word cloud of yeah at a prestigious academic center in the country in the south but it's representative of anywhere else in this country they asked what give us two words that describe your encounter with your doctor and it isn't very pretty as you can see and this is a recent couple weeks ago really great editorial by Daniel Alfre who's an author from New York City a physician author and she says if it requires thinking I'm sunk because there's no time and talks about all the tests that are ordered unnecessarily and added referrals because there's not enough time and finally she concludes beyond the financial waste modern medicine practical practice is a petri dish for medical error patient harm and physical burnout you may know this but physical burnout is a physician burnout excuse me is now an epidemic more than half of physicians suffer from burnout and 20% from clinical depression the record number of suicides in history of the medical profession and not just in this country but in most developed countries this is a recent magazine cover which is death by 10,000 clicks because as it turns out doctors now have become data clerks and so they not only are trying to deal with all this in in a very compressed period of time but they don't even see the patient literally because there's no eye contact because they're typing and I keep and this is the different specialties of Medicine they kind of all look the same don't they because that's what's happened now and it's a fiasco so I actually have a clinic every week and my main goal in my clinic I thought I should change it after Michael Lewis's book from the undoing project to the undoing clinic because I mainly spend time getting people to not to get off the prescriptions or to not have the tests that they otherwise were ordered because so much waste that's happening today because of this problem and just to give you a little bit more grounding on that the United States is a very negative outlier we are the only country in the history of the world since it's been recorded that's how decreasing life expectancy for three years in a row we also have the highest per person expenditures for healthcare $11,000 plus and we also have the worst infant child and maternal mortality of neoc ECD country so this is really pretty darn shameful and in addition to that we just keep hiring more people in health care so these are the data that show that as of late December 17 just a year and a half ago for the first time in American history there are more health care jobs and even retail so you need to keep hiring more people to have this model that doesn't work so what we're going to talk about now is a whole new model which is understanding each individual the essence of the medical side of each person and that was the subject of a wonderful National Geographic January issue if you hadn't seen it everybody is unique and we have the ways to understand each person now and it's many layers just like a Google map of a satellite view traffic view street view now we have all these different layers of a person that as far as their not just their anatomy through scans or their physiology through sensors all the different biologic layers DNA proteins metabolites the microbiome and even the quantification of our environment our exposome via sensors that can be done now and beyond all that there's the immuno which this group in Israel published a recent remarkable paper about how that status of our immune system is so predictive of what's going to happen throughout our life so just last week the Kelly twins Scott Kelly up in space for over a year as compared to mark the most detailed assessment of two human beings ever virtually every layer that I just mentioned for these two and it didn't bode well for anybody's interested going to Mars there was a very serious cognitive hit to Scott and also lots of things happening in his DNA and chromosomes suggest and he have just a year's worth of a flight would be a real risk for cancer now it isn't enough to assess this once it turns out beyond the DNA sequence which is fine to do one time a lot of the other omec assessments need to be done on a longitudinal basis like the immuno as an example so when I did this review article a few years back I called it from pre womb to tomb and one of my former mentors told me Eric you really had the wrong name for that article it should have been from lust to dust I said you're right a bill that would have had a much bigger up cake but as it turns out right now we are seeing finally medical sequencing not in the research silo but it's actually happening in the clinic it's happening here in San Diego at Rady Children's Hospital one of our partners in our efforts where they are sequencing sick newborns and doing that very quickly in fact within well within 24 hours and now in children with various conditions in in young and the young adults and even older adults who have unknown can serious medical conditions that can be diagnosed at least 1/4 or 1/2 just by sequencing to be able to determine the cause of a sis infections that are serious about not waiting for days for a culture to come back cancer genomic disease and we have a big program here in San Diego for molecular autopsy for sudden death and people unintuitive where the autopsy is negative and we work with the medical examiner to try to find for the family so they don't have to live the rest of their lives worrying about if they have a mutation that could put them at risk so that's a big program here at Scripps as well so how many of you have had a polygenic risk or anybody had that two people three okay good well that's a good thing reason you're here this evening so this is MIT tech review which is one of the better sources of information in this area that we're talking about tonight and they had their breakthrough technologies last year the number one was a polygenic risk score and about forecasting your fate not really fate probability of getting a condition by hundreds of differences in letters in your genome which you can get through a chip a 1 M 1 million chip which is run by places like 23andme or ancestry DNA and could theoretically be done for less than $20 and someday it will be hopefully in the near future so a liter Kimani in our group wrote a wonderful review article of very interested in this topic in last year and it's this whole idea that common conditions like heart disease and others that I'll mention hundreds of different letter changes in our genome indicate the risk and you can get a score and that's what this is about so these conditions like not just heart disease atrial fibrillation diabetes breast cancer prostate cancer inflammatory bowel disease they all have scores and this is a recent article in JAMA saying that these scores for common complex disease will become part of clinical care in the near future I would submit it should be today but we're not really ready for light to get people there their information so to me this is an app that we prepared Ali talked money and the team here at our Institute called my gene rank and while you're here right now if you had a 20-3 of me you can upload it to the app and you can get your heart disease risk score and hopefully it'll be real low so that is a step in the right direction it's free you are agreeing to participate in a research program if you do it and just today in sell one of the top biomedical journals we published Ali and I the obesity risk score the editorial about that paper which is it's harder to know what to do with that but it's something you can see in children when they're very young and the question is if you knew that that the child was destined to have obesity later in life could we intervene is there a way to do that which is palatable no one has the answer to that but the risk now can be defined and it's really quite extraordinary now let's switch from genomics which is not clearly not gotten out there in a really big way yet but what's just starting now to sensors which really are much easier to adapt to because we all have smartphones these days and there are sensors for every part of our body every system in our body one of the deep learning algorithms that was first the first one approved by FDA for for patients and the public was the atrial fibrillation detection so what it does is it picks up if your heart rate is normal at rest or with activity for you and gives you a signal that you should record your cardiogram through your watch if it's off track for you and so this can detect very accurately atrial fibrillation and may be useful in certain people but beyond that the idea is that you can do the same sort of thing for people with kidney disease who are at risk of developing very high potassium you can get without any blood a very accurate potassium which is pretty remarkable and that is giving you a sense that machines can see things that humans can by training with millions of data points and that's what something human vision could never see from a cartogram whether the potassium is four point three or four point four so this is a blood pressure watch that's now FDA approved we've been testing it and this is a really interesting because it gets you blood pressures at the least likely times like when you're in traffic or it never occurs in San Diego I know or if you're having an argument with your spouse that doesn't occur either I'm sure but I've been testing this it's a remarkably accurate and so we are moving towards being able to have blood pressure much more ideally even without having a press start in the times ahead glucose sensors have really zoomed forward whereby a glucose sensor it doesn't even need a finger stick calibration anymore so it's factory calibrated and you get a glucose every five minutes and this is really going to turn out be on your phone or your watch to be very helpful as I'll talk about in a little bit about diet and artificial intelligence and now connected sensors in the city of Louisville Kentucky they gave everybody a connected inhaler who had asthma just by doing that they reduce asthma attacks in the entire city of Louisville by fifty percent over the course of a year and they reduced as the need of inhalers by 70% so it just shows you a remarkable impact of people knowing where hot spots are where air quality issues are in one city now my favorite apple all is the smartphone ultrasound and this is remarkable because you just pop this onto your smartphone this probe and you can start imaging any part of your body except your brain okay and I think it's a remarkable advance it isn't being used in this country that much tied to issues like reimbursement but it is really a stunner because you get images that are exquisite as we've published that are as good as the three hundred and fifty thousand dollar machine that sits in clinics and hospitals this is what it looks like the heart that's my area of interest and you see this is typical images of the heart you see every aspect that chamber size the strength of the heart muscle the valves everything in second why would you listen to lub dub anymore when you can see everything and you share with the pitch and you can send the patient the video loops and it's a whole different world which is remarkable you can even track the leaks of the valves with this so I haven't used a stethoscope in almost a decade why would you do that when you have this so I was so enamored by this I think you would probably do something very similar if you had this in your hand I decided to a total body medical selfie so I I did my carotid my sinuses thyroid lung heart the liver kidney by the way I didn't know how to do these I only didn't to do the heart so I had it google it to find out how to do it which is easy by the gallbladder iliac artery spleen a Horta inferior vena cava popliteal fossa and my left foot now what's really interesting about this is I didn't know it would pan out I thought I did it as kind of a just a educational experience and then a few months later I had some abdominal pain I didn't know what it was so I was scanning my abdomen and I found my kidney was dilated my left kidney so I went to the emergency room here and I told the emergency room doctor I have a dilated left kidney on my smartphone here and he thought I was an alien anyway he didn't believe it so he sent me for a cat scan which is shown here and if you look closely at the images particularly if it wasn't so light up here you would see that this very expensive image here is identical to the one on my smartphone and it was free on my fart smartphone but I didn't believe it anyway I don't recommend this this is not a pleasant experience to Pasi stoned but it did show that the amazing capabilities of having a smartphone ultrasound that is gathering information this someday will be something that consumers will widely have access to now this week in Tuesday's New York Times was an article even though this smartphone ultrasound isn't being used that much in the u.s. here it is in Africa it's used widely in rural Africa hinterlands and it's diagnosing everything like pneumonia and you see here I mean you name it it's an incredible article if you want to take a look at it if you didn't see it in this tuesday's in new york time now beyond that capability now disempowerment like I showed you with heart rhythm is going to urinary tract infections one of the most common reasons to go to a doctor not in the u.s. of course but in the UK and so these are the various companies that have this and this is using artificial intelligence and the first one healthy il is actually in contract with the NHS in the UK and so you can see how it works I had needed this yeah it was quit but the whole idea is that this was trained by hundreds of thousands of people who had urinary tract infection so it's highly accurate and that's gonna be for many people the way to bypass having to go go to a doctor get a urine culture and all sorts of other things so just to leave you with the sense about the smartphone that it's becoming the hub of the future of medicine you're seeing the beginning of where this can go and this is going to be really important eventually for those of you who wanted to be a virtual health coach all the time helping you to prevent illness that you might otherwise get or manage ones that you have so I wanted to talk about where we're implementing this and Jamie mentioned this large grant that we have this all of us program this is run here at Scripps by Katie Baca moats in our group and this is an amazing program you're you're certainly welcome to join we have a hundred and eighty thousand of the million people in the US that have registered more than half are underrepresented minorities which is just in itself historical no no research program in the world has ever had that much representation by minorities and this is the website if you want to join join all of us org and we'd certainly the idea would be all these things that I've talked about will eventually be participants choice if they want to have their genome sequence if they want to have their gut microbiome assessed they want to use these sensors and so we're testing a lot of these things to eventually distribute them now let me get to the main topic which is artificial intelligence these neural nets which is what is the excitement in medicine this is from an update by Thomas Friedman in the New York Times in January he declared that the word for 2019 is deep I said well that's great that's the name of my book it's great and he was right it's really is true I wrote this high-performance review in Nature Medicine of course the book it is about going deep is using this data deeply understanding each human being using deep learning to move that data into an understandable and predictive format and then to restore this deep empathy so it started about 10 years ago so the Turing award which is a Nobel Prize in computer science was just awarded a little of a week ago to Geoffrey Hinton and his colleagues from the University of Toronto because they took images like cat videos like catheters but not cat videos in this case and they were able to use deep learning neural nets to get accurate depiction of these images and so now that's being used at everything you can imagine in healthcare the concept is this there's inputs it could be an image it could be a slide or an eye scan whatever you electrocardiogram it could be speech it could be text it goes through all these layers of artificial neurons and those layers the number of which are not dictated by a human a programmer which by the data itself and then there's an output which as this gets really good because these neural Nets have insatiable appetite for data unlike humans that have early satiety so this gets good and then it gets really super accurate as you'll see so the leading discipline is actually ophthalmology a lot of people think it's radiology but it's actually ophthalmology and so the first thing that really is shocking is if you show this eye retina picture to international retina experts and you say is it from a man or a woman and the answer is 50% right 50 personnel now there are other ways to tell whether it's a man or a woman I know that but but 97% accurate with the neural net 90 and no one knows why because it's been fed a million retinal pictures would the right answer as far as gender that's again to show you seeing things that humans can't see right training for example in Moorfields Eye Institute the leading eye institute of the world in London where by the way they do optical coherence tomography and everyone for their annual eye exam we don't even do that here except in rare occasions a no CTS are the best way to see the retina and diagnose a whole host of eye conditions and we don't even do those routinely so for people who think that we're so advanced here and eye conditions are really important and they were able to show with an algorithm a neural net algorithm deep learning that they could diagnose over 50 different eye conditions more accurately than retinal experts and other leading ophthalmologists so I'm not going to go through all these but every single clinician every type will be affected because of this capability and just to give you some other examples radiologists did you know that over 30 percent of scans are false negatives 32 percent that means they miss something important okay that's pretty bad and that explains why there's so much litigation and radiologists are expected to read so many scans no wonder there's a problem and now they're seeing AI companies that have been approved for algorithm support of radiologists in the u.s. here's an example lung nodules that could be cancerous are missed a lot and now they're not really missed when you have artificial intelligence to find them and then oversight by radiologists here's risk fractures which are frequently missed or misdiagnosed in the emergency room setting this can help merge your doctors become superior risk fracture diagnosticians it's not just radiologists here's pathologists can you imagine this that the pathologists can't tell what mutation it is by looking at the slide that's driving the cancer but by training with algorithms that can diagnose what is the mutation besides whether it's cancerous or not it's pretty striking again the human division limits versus machine vision training and what about change a cardiogram that you could tell the heart function we never thought we could do this from a 12-lead cartogram now we can from Mayo Clinic work detect what is the function of the heart skin lesions why would you go to a dermatologist when you can get the answer better than a primary care doctor or dermatologist by taking a picture and having a deep neural net applied to that picture because now compared to hundreds of dermatology experts the results were as good or better with neural nets and then this is colonoscopy this is of course everyone's favorite procedure and what's interesting there is that so many polyps are missed so many and now they're not going to be missed because deep learning of the video real-time randomized studies both of which are from China have shown that you won't miss these polyps so it's pretty striking but it isn't perfect we have seen for example a study of congenital cataracts which are difficult to diagnose we're trained ophthalmologists are better than the algorithms so far so it isn't I don't want to give you this sense that this is all a buttoned-up but we do have to test these things in clinical trials and then keep this under tight surveillance because if an algorithm has a glitch if there's malware or if it's being used in a different population from which it has been tested and validated it could create lots of potential harm now I had the opportunity speaking of the UK of being commissioned by that government to run a review a planning for the next 20 years using AI and this is a really exciting opportunity to learn about the NHS which by the way unlike the US the NHS is regarded above the royal family and the BBC as the most revered institution in the country not like what we think of health care in the US right so we did a review and we ranked all the different technologies that are going to have an impact ai is the most transformative of all and I think it's really clear that they are planning ahead we are not doing any planning we had an announcement earlier this year of the American AI initiative zero dollars were allocated zero and we have no planning in this country for the use of AI in anything no less than healthcare so now let's turn to the clinic we talked about the doctor that's looking at the keyboard and the patient that is saying what am I doing here right well now it turns out there's over 20 companies that are using another form of AI called natural language processing NLP to synthesize the note to liberate from keyboards and much better notes by the way it already implemented in the UK hopefully soon here in the US to get rid of human scribes there's over 50,000 human scribes now in this country that are there so the doctor can look at the patient it's amazing so we have to do better than that that's like going back to before Gutenberg you know that's really kind of difficult so this is a most famous picture of painting ever in the history of medicine it's called the doctor by Lucas Lucas and this is a modern depiction of that here you see we want to move away from this okay we want to get restoration of caring for patients now in the UK the reason I mention it is they economic economist scuse me were involved in our multidisciplinary review they found that even just one minute of liberation from keyboard it was one of the most productive things you could ever have in healthcare it equated to enormous number of increase in doctor slots to see patients and care for patients so this is an effect that's not only having all clinicians but across all the health span everything from before and embryo is implanted all the way through one's life at the stage of potential end of life and just to reinforce that this was published just last week and it's a study which is pretty remarkable because so many embryos are discarded that actually are of high quality but it's because the embryologists that's advising the couple can't discriminate because it's so subjective now using machine vision the number of embryos that are of good or high quality is markedly augmented that's really exciting because you know the problems we have with in-vitro fertilization and successful term pregnancy I want to show you this about the patient status today so that that's the beginning of the movement and when you have data that's eminently portable and you have people with high-speed Internet connectivity there's no need to have paternalism and the patient will see you now is so well graphically shown there so for patients what is a guy going to do for patients I already mentioned things like a heart rhythm and UTI but I thought you really really interested in diet because a lot of people would like to have their diet be part of their medicine they're promoting their health food is medicine when that'd be nice instead of having to take prescriptions so this was a time feature about this whole concept but recently I wrote an essay for the New York Times if you saw it called I didn't call it the AI diet they did but that's essentially we cracked the case finally with a honor and it wasn't we it was actually the group in Israel at the Weizmann Institute a few years ago the point being is up until now you've seen all these guidelines about what you should eat in food pyramids the same diet for everyone is totally wrong totally wrong because we each are unique why would we eat the same diet so finally the group in Israel cracked this case and they published a paper in cell they published a book called the personalized diet and the point being is all of us ate the exact same food the exact same amount the exact same time our glucose after we ate would vary all over the map and that has been replicated by many other groups and now it's been extended to your triglycerides in your blood so we have a unique response to food finally we have nailed that and the way it was named was they gave thousands of people in Israel these controlled food diets they measured everything including Kouzes and their gut microbiome and all the different layers of information about these each person and they showed that there was this highly individualized response and much of it not all but much of it was actually influenced by your gut microbiome so they accompany a datura was Siri what should I eat and I'm not sure that we would want to rely on Siri which is not a very advanced form of NLP but well you get the you get the message so I decided I thought this was really cool I want to try it so I wrote to the fellows who I'd met in Israel and I asked him if I could get a kit and be one of their participants so I did and I had a glucose sensor and I had a lot of spikes and I don't have diabetes and in fact a lot of you if you did this would have a lot of spikes to certain foods that you would never have predicted I also had a midgut microbiome and I had to type into my phone everything I ate and drank and every sleep and physical activity and medications it was a real pain in the neck to do this for two weeks and it isn't available this package is not commercially available there's scaled-down things that I don't believe really work and you'll see this might not even work either but it was an interesting experience because it gave me output of everything I should or shouldn't eat for me to avoid glucose spikes and one could argue well maybe glucose spikes are not the be-all and end-all end point but nonetheless my favorite foods got a really poor rating and the things I would never go near like bratwurst cheese cake things that I thought were potentially lethal they got a plus ratings except this is incredible so anyway though we're heading that way where you will if you want have a diet that will tell you how to avoid glucose spikes how to avoid triglyceride spikes and many other things that are undesirable it'll take years before we have randomized trials to show that it prevents diseases like diabetes or heart disease but interim this is coming and you might be ready for you'll see these sort of things over the next cut for years and then lastly about the levels of influence I'll just say the Health System has big impact because you can do a remote monitoring for years for the cost of one night in the hospital in the US which is over five thousand dollar charge and so we have sensors that we can do that and we're gonna have all sorts of AI to replace hospital rooms not the ICU or not the emergency rooms or opportunities but the rooms in a hospital so just to kind of pull this together we are still at the early phase of taking in this multimodal data like sensors and genome genomics and your electronic record but that's where this is headed and what that gets us to is a not an autonomous form of healthcare for certain things like a ear infection or a skin lesion yes but for most things it still will require human doctor clinician oversight so unlike a driverless cars which will never be fully autonomous we're not even going to get to the point where they are because here anyone with a serious diagnosis would like to have a human contact human oversight to make sure that it's right and what are we going to do about this and also to learn about the person's story so learning about each person deep learning is going to look like this this virtual coach is going to have all of your data if you want to sign up for this it's also going to have the corpus of the medical literature up to the moment relevant to you and it's going to have an avatar talking to you one way or another to give you advice and in the book that you have you'll see many examples of what that's going to look like in the near future it's pretty exciting because this is all possible and if you if you talk to the AI gurus around the world they're very excited about what this can do for the future of healthcare so basically then when you have doctors that have less tasks less or no data clerk function more time and you have patients that are taking on more charge you have a flywheel effect where you can actually see the restoration of the case in healthcare which is so exciting which we've lost there's been serious erosion over the last at least three or four decades and that is the most exciting part of this I don't want to leave you with the sense that there aren't serious issues that have to be addressed there's a chapter deep liabilities in the book that addresses all these serious liabilities my biggest concern is not about these they're all important my biggest concern of whether this great jump in productivity efficiency and workflow and accuracy and speed which would give the gift of time would it be used to make things worse that would be the default mode if we don't stand up and say we need to give this back to patients give this back to the relationship the human bond the human touch the deep empathy so this is just Abraham Verghese is the leading medical humanist of our era he wrote the foreword of the book and he has in that forward a quote about intok from the book intoxicated by by illness and he recalls this one part about my soul as well as my flesh to get at my illness for each man is ill in its own way the point being is we don't we'll never digitize a person's life story that's the essence of medicine that requires time and that's what we want to get back to and those of you are old enough to remember 40 years ago when I finished medical school that was routine but we've lost that we've lost the mission for doctors we need to restore that because they don't want to be data clerks they don't want to be saddled with all the little time and little ability to provide care so what we need to do today human performance is not going to change it may vacillate a little bit but over the times ahead it's pretty stable and what we want to do we know machines are going to keep getting better that insatiable appetite for for data and algorithmic training they're gonna get better on so many tasks we are just seeing the beginning of that now what we need to do is get more human more you mean and if we do that in healthcare we can make a really big difference it's really exciting so I just leave you with this get back to the deep stuff deep understanding of each person deep learning to deal with that data and the restoration of this deep connection and empathy and respect restoration of way medicine used to be these are all the people I get to work with a real privilege in our group and being working at Scripps research is is a dream to be able to have be creative and innovative among an extraordinary community of faculty and grad students and everybody here it's just an amazing place and no wonder it was ranked number one by nature for innovation place so I'll leave you with that open it up to questions thanks so much for your attention [Applause] you

Leave a Reply

(*) Required, Your email will not be published