Analyzing speech to detect cognitive impairment associated with dementia and mental illness
Recommended Book:
Best Way to Contact Liam:
liam@wintelightlabs.com
Company Website:
Hey Outcomes Rocket listeners, thanks for tuning into the podcast again. Tired of your businesses health care costs unpredictably increasing every year? Health care costs are typically a business’s second or third line item expense. And if you’re like most employers, it’s an expense that’s growing faster than your revenue. Luckily for employers Noveta Health has the solution. Noveta Health is a full service health care consulting firm with proven strategies to lower your health care costs by up to 30% or more. They operate on a fee for service model and never mark up any of their medical or pharmaceutical claims. None of your employees have to leave their doctor or pharmacist either. Their health captive and pharmacy benefit manager are the most cost effective and transparent solutions in the whole country. What they do is not magic. It’s just honest. So if you’re tired of over spending on health insurance and want to learn more, visit outcomesrocket.health/save for a free spend analysis to see how you too could save by switching to Nevada health. That’s outcomesrocket.health/save for your free spend analysis outcomesrocket.health/save.
Saul Marquez:
Welcome back to the podcast today I have the privilege of hosting Liam Kaufman. He’s the CEO and co-founder at Winterlight labs. They are commercializing a proprietary language-based diagnostic system that analyzes natural speech to direct monitor dementia, Alzheimer’s, aphasia and various other cognitive conditions. Winter lights scalable platform uses short recorded speech samples to analyze hundreds of linguistic cues and can detect dementia and other conditions with a high level of accuracy. This is a major improvement over our current pencil and paper tests which are time consuming costly and difficult to administer. The platform has applications and drug trials, long term in primary care and speech language pathology. Liam has published peer reviewed articles in Cognitive Neurology, human computer interaction and neuroscience. He’s an experienced software developer and has successfully launched understood at a startup acquired by EventMobi and featured in the Toronto Star CTV News and TechCrunch doing a lot in the space to improve outcomes and that’s why I wanted to get him on the podcast today So Liam want to give you a warm welcome thanks for joining us.
Liam Kaufman:
Thanks for having me.
Saul Marquez:
Absolutely. Now did I leave anything out in the intro that you want to share with listeners?
Liam Kaufman:
No, I think that was pretty comprehensive.
Saul Marquez:
Awesome. Glad to hear. So what is it that got you into the medical sector?
Liam Kaufman:
Yeah for sure some are really interested in this intersection between medical science and computer science. I think at this point we’re barely scratching the surface in what technology can do for medicine what it can do for patients and so I think over the next couple of years and over the next decade we’ll see huge impact on care. And it’s just excited to be a part of that.
Saul Marquez:
That’s really great yeah. You know there’s there’s definitely a lot going on lately and the improvements are exciting. If you had to call out anything in particular, Liam, what would you say is a hot topic that needs to be on health leaders agenda today and how are you guys approaching it?
Liam Kaufman:
Sure. I mean I’m obviously pretty biased but I’d say A.I. is a huge topic. I think you know there’s a lot of hype around A.I. and so I sort of see two extremes in the industry I see people who are really very skeptical of new companies touting A.I. and then I also see people that are you know sort of blindly excited about using A.I. and I think ideally we would be somewhere in the middle where people of a healthy level skepticism. But at the same time they educate themselves and understand just how A.I. works how can be plugged into their current organization and how it can benefit the patient.
Saul Marquez:
And so as we dive into A.I. in particular you and your team are interested in voice and so tell us a little bit more about what you guys are doing in that space. Maybe some examples of how you guys are improving the outcomes needle on that.
Liam Kaufman:
Yeah for sure. So you know right now the way that cognitive impairment is assessed or even psychiatric illness is often through either subjective assessments or incredibly time consuming assessments or very expensive neuroimaging. So I think there is a lot of opportunity here to improve the diagnostic process and so our observation was that the voice is really important for conveying cognition and mental health. I mean that’s sort of like informal basis. We know when someone sounds sort of tired or you know someone sounds sick or confused and what we can do is we can actually use A.I. to pick up on all those little markers and we can use AI and algorithms to actually measure these markers objectively. And so to take sort of a concrete example. Someone who has mild cognitive impairment that precedes Alzheimer’s disease they’ll start to speak differently. Five to ten years prior to diagnosis and so what they’ll start to do is they’ll start to forget people’s names so they’ll use fewer nouns and they’ll start to use more pronouns like he or she etc. will use words that are less complicated instead of saying the word SUV or sedan they might just say the word car because it’s easier to remember. So we can objectively quantify those observations then we can use a A.I. to determine if they might be developing a condition or not but we can also sort of measure the severity of the impairment using AI.
Saul Marquez:
That’s pretty interesting just taking a look at some of those trends and and what could be indicative of cognitive impairment. Super fascinating to learn and know that these things are being put into a computer and into an algorithm. Tell us a little bit about potentially one of your algorithms or an exercise that maybe didn’t work out as well and some of the learnings that you came out with that made you better.
Liam Kaufman:
Sure. I think one instance was more on the business development side. So when we first started commercializing the technology in 2015 we had a number of different thoughts on what or which verticals could benefit from this type of technology. So no one might be senior care another might be pharmaceutical companies and so we had lined up this pitch with someone from a senior care company. And so I spent 40 minutes walking her through the technology how it works et cetera. And then at the very end of the pitch she said OK why am I here. And then there is you know silence in the room and afterwards I started to realize that you know this is probably pretty obvious to most of your listeners but you know when you pitch someone on a new technology you really have to personalize that pitch so that you sort of get why you’re talking to them you know how that technology can fit into their industry how can help them. And a week later I actually had a pitch with someone in pharma and because that pitch with the person that senior care failed so miserably I spent quite a bit of time personalizing the pitch for that Pharma partner and twenty five months later we actually closed the pharma company as our first customer. So I think for me it was sort of a failure in how I communicate with the technology and so are how I sort of educated that potential customer on the benefits of software.
Saul Marquez:
That’s a great watch out and thanks for sharing that. You know we can always appreciate that the importance of doing that you think that your customer knows exactly what you do it’s sort of like the curse of knowledge right. You put them on the same platform that you are when really they’re more like a one or two on a scale of 10. So I think that this is a great learning, Liam and you obviously found success when you corrected the way that you did things. Now you guys have been at it for a while. You got customers things are cranking. What would you say is one of your proudest leadership experiences?
Liam Kaufman:
That’s I think one of the proudest moments that we had was we had a very skeptical customer last year and I allude to this earlier in terms of a I where you have people that are extremely skeptical or not skeptical at all. This customer was definitely in sort of the sceptical end of the spectrum. So they really want to test our technology and sort of the most difficult way and so the way that they did it is they provided a blinded data set. I can’t go into sort of the specifics but it was a much more difficult task than we had ever done before it was blinded. So we didn’t know who had mild cognitive impairment and who didn’t and there were some other parameters that made it a really challenging task. And so you know we weren’t even sure if we wanted to do it just because it was a little bit outside of our wheelhouse. We ended up doing it and we ended up beating the baseline for that type of diagnostic by 25% and we ended up signing a second deal. So that was we’re really sort of exciting that we went from a point of sort of extreme skepticism. We overcame it and we ended up signing a second deal.
Saul Marquez:
That’s awesome. Congratulations. Thank you. It’s not easy to do and especially when it’s out of your wheelhouse and you’re having to push push it. So what was the deciding factor there that made you say you know what I’m going to do it versus you know what this is not our area of expertise?
Liam Kaufman:
I think it was close enough to our area of expertise and I’m an optimist and I think you have to be when you build a new business and so I firmly believe that we could do it. So we ended up doing it and we ended up succeeding.
Saul Marquez:
Nice. So what about today. And what’s an exciting project your focus here you’re working on?
Liam Kaufman:
Yes so we’re thinking about how we can help patients in a much broader context outside of pharma. And so one of the areas that we’re pretty excited about is more specifically Medicare Advantage. And so what we’re saying is that there is a whole bunch of folks that have chronic health conditions and there are different things that impact their health outcomes and couple of things that impact how well they recover or how well they do in the long term are things like cognitive impairment and depression. So if someone has a bit of cognitive impairment they’re going to forget to take their meds, they’re going to have to go to the doctor’s appointment or monitor their blood sugar as effectively if they’re depressed they might be apathetic. So there’s all these sort of mental health things that impacts someone’s health. And what we’re saying is that we can help identify those people much quicker. We could do that remotely and then so we can help Medicare Advantage plans pinpoint who could benefit from extra care whether that extra care the home care or extra physio etc.. And so we’re quite excited by that.
Saul Marquez:
A very simple way to get ahead of potential problems.
Liam Kaufman:
In fact.
Saul Marquez:
And at what point of the interaction with the person on Medicare Advantage would you guys be inserted.
Liam Kaufman:
I think there’s several points. I think the most obvious is sort of the yearly wellness visit. I think that’s a natural point. Many of these plans also have call centers where they check up on people in their plans. And so you know technology doesn’t just work over iPhones right but they can also work over telephone call. So that could be something that patients could often too. So is a lot of different touch points and sort of exploring those right now.
Saul Marquez:
Very cool very cool definitely exciting and so thinking through what opportunity we have in this space listeners is paying attention to companies like Liam’s Winter Light labs taking these these initiatives to the next level to make them real. Voice is definitely a powerful biomarker and one that we should consider. And so getting close to the end of our time here with Liam this part of the podcast we’re going to do a syllabus. So I’m going to ask a couple of questions in a lightning round and then we’ll follow that with a book you recommend to the listeners, you ready?
Liam Kaufman:
Yup
Saul Marquez:
All right. What’s the best way to improve health care?
Liam Kaufman:
I think the best way is really making sure that incentives are aligned with the best interests of the patient.
Saul Marquez:
What’s the biggest mistake or pitfall to avoid?
Liam Kaufman:
Business development in health care takes a long time and so I recommend not pivoting too soon.
Saul Marquez:
Great advice. How do you stay relevant despite constant change?
Liam Kaufman:
Really it’s about reading a lot. Listening to customers clinicians and patients.
Saul Marquez:
What’s one area of focus that drives everything in your organization?
Liam Kaufman:
Ultimately it’s just about improving outcomes whether it’s patient outcomes or clinical trial outcomes. That’s our focus.
Saul Marquez:
Love it. These next two are a little more on a personal note. What’s your number one health habit?
Liam Kaufman:
So I’m a huge fan of exercise and running and I personally love running. It’s so portable and so even if you travel frequently you can still run whether you’re in Boston, Chicago or Toronto and it works sort of any time of the year.
Saul Marquez:
That’s awesome. And how about your number one success habit?
Liam Kaufman:
It’s forming habits like just sort of in general is really important. Doing things on a regular basis removes that level of difficulty if you just do it frequently, it doesn’t. You don’t think of it as being difficult anymore.
Saul Marquez:
Love it. What book would you recommend, Liam?
Liam Kaufman:
So this might be a bit controversial but I think so too in the last year one of the books that I really enjoyed was Bad Blood. It was credible page trainer but I think our industry often focuses on success and I think what I liked about this book is I mean I was really engaging but it also sort of highlighted incredibly dishonest company that failed. And I think there’s a lot of there’s a lot to learn around that story. It’s an excellent cautionary tale obviously for investors but also for employees for founders in the health space that creating a business from scratch is really difficult. Creating a business in health care is probably in order of magnitude more difficult and I think even though this book is focused on there and also I think it highlights a lot of the challenges that you face in health care whether it’s pivoting to different business models regulatory and I think it sort of highlights a lot of these issues in a more engaging way. So I mean ultimately I think I would recommend the book for your listeners.
Saul Marquez:
Great, great recommendation there, Liam. Folks, you could find that recommendation as well as an entire transcript to today’s interview. Just go to outcomesrocket.health and in the search bar type in Liam Kaufman or you could type in Winterlight labs you’ll find all of the resources we discussed there today. So Liam I love if you could just leave us with the closing thought and then the best place for the listeners could get in touch with you.
Liam Kaufman:
Sure so now seeing tremendous change in pharma over just the last 12 months there’s a lot of interest in digital biomarkers and digital endpoints. So I think we’re gonna see a lot of announcements this year but especially in 2020 as one of these technologies become used in clinical trials and I can be reached at liam@winterlightlabs.com.
Saul Marquez:
Outstanding Liam. Hey, really appreciate you spending time with us today and excited to see the next step you guys take in the development of this.
Liam Kaufman:
Thanks for having me.
Thanks for listening to the Outcomes Rocket podcast. Be sure to visit us on the web at www.outcomesrocket.health for the shownotes, resources, inspiration and so much more.
Sonix accurately transcribed the audio file, “Assessing Cognitive Impairment with Voice Markers with Liam Kaufman, CEO & Cofounder, Winterlight Labs” , using cutting-edge AI. Get a near-perfect transcript in minutes, not hours or days when you use Sonix. Sonix is the industry-leading audio-to-text converter. Signing up for a free trial is easy.
For audio files (such as “Assessing Cognitive Impairment with Voice Markers with Liam Kaufman, CEO & Cofounder, Winterlight Labs”), thousands of researchers and podcasters use Sonix to automatically transcribe mp3 their audio files. Easily convert your mp3 file to text or docx to make your media content more accessible to listeners.
Researching what is “the best audio transcription software” can be a little overwhelming. There are a lot of different solutions. If you are looking for a great way to convert mp3 to text , we think that you should try Sonix. They use the latest AI technology to transcribe your audio and are one my favorite pieces of online software.
