Sharief Taraman Presents Cognoa at LSI USA ‘23

Transcription

Sharief Taraman  0:03  

So hello, I'm Sharief Taraman. I'm a pediatric neurologist, as mentioned, and we're going to be talking about children health. If you don't like children, you're in the wrong room. But that's what we're going to do. And why do we care about children? Well, we care about children, because there's many, many children like this young lady Charlie, and Charlie has a neurodevelopmental condition. Her mom didn't know what it was. But she knew that her daughter was lining up toys, she knew that her daughter would get upset when she moved her toys out of the line. And that was different. That was a typical, it wasn't what she expected. Now, as I mentioned, I'm a pediatric neurologist. I'm one of about 1500 in the entire country. And I love what I do. I love taking care of kids. But I have a problem. I've got too many kids to take care of. There's 750,000 kids here in this county alone. And when I started as a division chief at the Children's Hospital, I had 1000 kids on my waitlist. That was 10 years ago, and I still have 1000 kids on my waitlist, because we can't see them there can't recruit enough pediatric neurologists. So how do we solve the problem? Well, parents drive, they drive hours to get their kids to a specialist to see somebody. And then when they get to that specialist appointment, they undergo comprehensive evaluations would take hours of time to complete. And many of these centers have long waitlist waitlist of over a year to get an evaluation. And what happens is you miss this early neurodevelopmental window, you miss the critical window to effectively change these children's trajectory. Parents know something's wrong in 14 months, but they can't get in to see the doctor to get a diagnosis and get treatment started, on average, 4.2 years, it takes them to get a diagnosis. That's too long, we've missed the early winner window of opportunity. And if you get these kids in that early window of opportunity, what you can see is you can actually improve their outcomes, so much so that they no longer meet criteria for autism. And that's the beauty of being neurodivergent, but not actually having to disorder. So how do we solve this problem? Well, at Cognoa, we are about access, we're about equity, we're about responsible AI, and we're about taking action. And so what we did is we developed the first of its kind, FDA authorized device to diagnose autism using artificial intelligence, it was a de novo application. And the device works extremely well. Before this, what we did is we used to take research tools, I would take them and put them into my clinical practice. And those research tools were normalized and based on predominantly white male children of high socioeconomic status. And so what happens across the board is you miss the kids. And now the beautiful thing about what we do is because we collect data from the parents, they download an application answer questions, they upload videos of their child in their normal environment. And then we collect data from the physician, but do it in a hyper efficient manner, that I can do it in about 10 minutes worth of my time, not hours of work in a clinical setting. And I capture all of this information, not a single biomarker, but many, many features of that child. And we leverage machine learning to understand what are the maximally predictive features of autism, but not just autism. I'm collecting a pipeline of children who have speech and language disorders, ADHD, and anxiety. And so we really fundamentally understand these children in a high dimensional space. And part of the problem in medicine, if you're trying to tackle any neural problem, any problem in medicine, really, we tend to lump things together. And when you lump things together, you lose clarity, you cannot understand what is the way that you can really target these patients. What how do you measure response to treatment? And how do you actually do genotype phenotype mapping? Because we have a high dimensional phenotype, we can actually start really understanding the phenotype as we move forward. So how well does the device work? Well, I'll tell you how it works. If you look at a comprehensive evaluation at a neuro clinic or developmental center, we know that 15% of individuals are going to be told you don't have autism go home. But in fact, they do have autism and they get missed. And if you look at their resounding number of children on that negative predictive value miss, there are misdiagnosed. It's females who are excluded from the early descriptions of autism. It's Hispanic, African American, low socioeconomic status, kids from rural counties. If you look at the positive predictive value today, just today, the CDC announced one in 36 children will be diagnosed with autism that is the prevalence, not one in 44, which is what it was before the last year and not one in 100, which was what it was 10 years ago. So the prevalence is increasing. Part of that prevalence is because we are doing a better job with the equity piece a little bit better, but not really. But also the false positive diagnosis rates pretty high. I think we're labeling children. I know kids are getting labeled with autism when they in fact don't because he Even after that comprehensive evaluation, what's happening is is that the competence of the clinician is only about 65%. Those were the bars that we had to be to get FDA approval and our de novo application. How do we do? Well, we come in at 96%. For our NPV, we come in at 88%, for our PPV. And the beauty of an AI device just like a clinician, if you go to a doctor, and they tell you that they never miss anything, and they know everything. I would not trust that Doctor, I have the courage to tell patients, I don't know, they present with weird stuff. Trust me on pediatric neurologists, I see lots of weird things. I have no idea what's going on with some of these patients, it happens. But we should hold the AI to the same accord. And so if you're looking at AI based tools, and they don't have the courage to abstain when they don't have an answer, I would be very wary of investing in those those companies. We are one of the very few that actually say we don't know. But we give value in the I don't know, because we can present the clinician with enough information, they can still take action. They may not get to the autism or not autism diagnosis, which we were aiming for. But they know what to do with the child and take the next steps that matter. Does anyone know where this is? So geography quiz, you didn't know I'm gonna quiz you. I'm a professor of pediatrics. Of course, I'm gonna quiz you. This is beautiful, Wyoming, beautiful state. But the last developmental pediatrician that they had in that state, retired in COVID. She was like, I'm done. Where are they going to send their kids, they're going to drive to Seattle Children's, or they're going to drive to Denver, Colorado. That's what they're forcing their patients to do. Until when, until we showed up. And we said, You know what, we'll give you a canvas, the x, we're going to work with the state of Wyoming. So we've launched in February in the state of Wyoming. And we're providing that service to patients in Wyoming so they can get access, and they don't need to take their kids out of state to get access. Medicare is huge, this huge population of children who get Medicare services, we've just inked a deal with perform RX that opens up 14 state Medicaid to us manage Medicaid systems in the United States where we're starting to commercialize. We also have agreements with self insured employers, including HTA. So that gets you the large self insured employers in the world. So we can offer it as a benefit for employees so that their families, their employees don't have to wait years to get a diagnosis where you have an absenteeism of an employee, they can get the answers that they need, so they can go to the next step. And then we working currently also with ESI, to capture the rest of our states. And we're working on that longish arc, that we all know, all right, your your journey doesn't end after FDA market authorization. That's where your journey starts for commercialization. And but we've made a lot of progress. And the beauty of the progress that we've made is that every state has a mandate that they must cover, autism diagnosis and treatment. And we're the only thing that's FDA market authorized. So every policy that doesn't updated their policy to include Canvas, the ex is actually violating their own legal statutes within their, their their state. Now, we've solved the autism diagnosis problem. I've got that covered. I took care of it. But how about the therapy? therapy's not as bad you can get in but it takes weeks. So is there something that I can do in those weeks? while they're waiting for therapy? Absolutely. We have breakthrough designation from the FDA for an autism therapeutic. Now the traditional therapeutics you can see her in the bottom right corner left corner is a therapist, and you know, she's using some flashcards here, this is happy, this is sad. Socialization is a fundamental challenge in autism spectrum disorder. And so what we can do is we can digitize that we can use AI, and we can give kids their normal environment where they can see faces recognize it. And in the in the few weeks of therapy that we delivered, that got us our breakthrough designation. We're able to do the same thing that would take 40 hours a week of a therapist delivering therapy for two years. I can do that in six weeks. So then the day do you want to fight for kids and help me do that? Thank you