Name: Brainstorm AI 2024: Improving Autism Care With Intelligence
Uploaded: 2024-12-10
Duration: 8 min 5 s
Description: Presenter: Ling Shao, Founder and CEO, SpectrumAi

Presenter: Ling Shao, Founder and CEO, SpectrumAi 

Brainstorm AI 2024: Improving Autism Care With Intelligence

Conferences

My name is Ling Shao. I'm the founder and CEO of Spectrum A I. And I'm here to tell you a story of how an autism start up found and stumbled into a really unique business model of how to collect labeled human behavior video data. Now, at our core, we're trying to help improve the quality of care for kids in the spectrum. But in the process, we managed to find a way where we can get paid to collect this labeled human behavior data that has implications far beyond autism and even far beyond health care and in other industries, for example, uh education. But let me start at the beginning of my story. So I'm a mom of four boys who are all on the autism spectrum. And for those of you who are not familiar with autism, it is uh in essence uh a condition where kids have really significant challenges, learning language and how to socialize with others. And even after they learn language, sometimes our kids are uh labeled as acting like robots because their uh their physical kind of how they portray themselves physically does not necessarily line up with how they want to communicate and sometimes they speak with a really flat intonation. Now, full disclosure, uh I went to MIT as a nuclear engineering undergrad. So I'm very familiar with being socially awkward, which is makes this a little bit really awkward right now. Um So I thought my kids were going to follow in my footsteps, but as it turns out, they had a real medical condition and um because they all started out nonverbal, the only treatment covered by insurance was called applied behavior analysis. And what that means at its core is basically you have clinicians using behavior data of how your kids respond to try to figure out how to teach them how they learn what's the right education protocol to apply to them. And now my little guys uh they all learn through what I call pattern recognition, which means that when we are teaching our youngest son, the word for ball, we showed him a picture of an orange plastic ball and said that's ball. But he could not generalize ball to a picture of a green ball or purple ball or even the 3d actual object of that orange plastic ball. We had to show him tons of different photos and even 3d images or 3d objects to say, what are the possible things that could correlate with this word? Does that sound like something else we train? Right? So I had this light bulb moment of oh my gosh, the way we train our kids, how to use language and socialize with others has a really close parallel with how we train our A I and ML models. But let's hold on to that light bulb moment because I had a more pressing need. I had to figure out how do I find the best clinicians with the best clinical protocols to treat my kids? And as it turns out, as you heard in the intro, I am not alone, the incidence of autism diagnosis has increased over 300% in the last 20 years. In the last five years, we now have mandatory insurance coverage in all 50 states. So I started Spectrum A I as a way to pay it forward and help all families like mine be able to access high quality clinical care. The way I was going to do that is to marry the insurance reimbursement to pay higher for those clinicians who actually deliver better outcomes. Now that sounds super obvious and super simple and for the healthcare folks room that's called die based care, but it is incredibly difficult to pull off because we don't have objective data that the health insurance companies can audit to prove if these kids are getting any better. So the way we solved that at Spectrum A I was we built an electronic medical record, the clinical documentation portion of it to enable therapists to collect data during their sessions. And more importantly, they were labeling the video of the actual sessions. So we've recorded the video of each session and as they were taking data on how the kids were behaving, they were actually labeling our video. Now, this is important for a couple of different reasons. One, it helps clinicians collect data easier and more time efficient. So they're spending more time with the kids and less time creating reports. But more importantly, it gave health insurance company an objective bank of data sources that they could use to compare across aggregate physicians. And then because they could audit the system, they can actually reimburse based on the outcomes the clinicians were actually delivering. So I'm really happy to say that three years into our start up this year, we had the first national health insurance company launch an initiative where they paid providers who used our software, a higher reimbursement rate because they had access to this objective data. And in Q one, we're actually the launch with two large Bay area employers. I'm sure you guys have all heard of um to allow their families to access higher quality providers as part of their employee benefits. So in the near term, I think it's a pretty easy jump for you guys to see how we're going to use this data to actually start to automate labeling of some of these more common behaviors, right? That again saves the clinicians more time. But in the longer term, as we collect a larger and larger data set of accurately labeled human behavior data because remember it's being labeled by clinicians who are literally trained to label human behavior data. The use cases of this data goes far beyond just improving autism care. So for example, education, imagine a world where we can measure and figure out which kids actually learn better using which teaching techniques. I started my career out as an army nurse and I can definitely see a military use case for this. Imagine the military using this to assess behavior of their troops as we think about discharging them for battlefield readiness. Now, I'm going to end with bragging a little bit about our tech because uh here is actually a footage of my youngest, remember the kid who couldn't generalize the word ball. Um This is actually a screenshot of the interface that the supervisors of the front line staff used to see the video that's been labeled during his session and allows them to both edit and annotate review and annotate what is actually being labeled so they can make the data source even more accurate. And two, this is the data that insurance companies can review and audit to see what progress the kids are actually making so they can validate why they're making these higher reimbursements. And if you zoom in, you can see that uh while we made this system to really help label the behaviors of these kids to figure out which protocols would help them best this label data that we have generated and get paid to generate is really what we think could be the best value of our company. And it's something that we stumbled into but believe have far-reaching applications beyond just helping these kids. So we started using this data to help kids who um acted like robots how to communicate and socialize with other, with their peers. But nothing stops us from using this data to literally teach actual robots how to interact with each of us. So for any of you who feel like this data could be used for something you're working on. I hope you will join me on my journey to use this data. Thank you.