Maria Artunduaga, Samay - Spotlight Interview | LSI USA ‘23

Transcription

Nick Talamantes  0:14  

Maria, thank you so much for joining me in the LSI studio today.

Maria Artunduaga  0:17  

Thank you for inviting us.

Nick Talamantes  0:18  

Tell me a little bit about the work you're doing at Sinai.

Maria Artunduaga  0:21  

The company was inspired by personal tragedy. This is a prototype called Sylvee after my grandmother, I lost her to COPD exacerbation. That was misdiagnosis a couple of years ago. And we what we are building, it's an AI enabled management platform for respiratory health anywhere.

Nick Talamantes  0:39  

How does it work?

Maria Artunduaga  0:40  

How does it work? So we have this prototype device, we are using active acoustic sensors. So what it means is that we are placing this prototype on somebody's chest. Right? Right. And we are sending signals all from one end, right here the speaker. And on the other end, we are listening to that sound that transmitted through the chest and we are catching resonances. What are the main principle of the idea is that when you have lungs that are functioning, functioning, okay, it has a particular resonance profile on wonder the lungs are declining, we can catch those changes. The main thing that we're trying to do is to improve pulmonary function when you are outside of a hospital. So by using a wearable all the time, we can know that without asking patients to blow through devices, which is one of the main reasons why we are losing people to COPD or asthma or lung.

Nick Talamantes  1:42  

Yeah, it's my understanding that there are millions of patients that have asthma and COPD together. And the main cost and risk to them is when they have an exacerbation.

Maria Artunduaga  1:54  

Yes, COPD are, just in the US just under US, so 30 million Americans have COPD, half of them are not diagnosed 26 million are affected with asthma and about 18 million have lung. The main issue with COPD specifically and let's talk about them, it's like 50% of exacerbate every year at least once and every exacerbation costs $20,000 just one single event, you go to the hospital, sometimes you are admitted to the ICU, it costs a lot of money. So the main thing is we're trying to replicate what other companies in digital health have already done. So they use wearable sensors, anything sensor ish at home that can figure out continuously remotely and easily how your organs are doing right. So you have a rhythm for cardiac stuff, you have continuous glucose monitoring for diabetes. By the way, I'm a user. I love it. I love the whole experience. So that's what we're trying to replicate with some AI right? So can we figure out as a COPD patient Your lungs are doing okay? Can we pick up signals that are showing us some sort of decline of the function of a lung so not symptoms is not cough or wheezing or anything like that actual function meaning both the air coming in and out there Weiwei, how he's expected. And then we can enable kirkko coordination with actual providers, it could be either us or an actual pulmonary clinic that is taking care of these patients. So the full exacerbation is, you know, scalable, you know, with that it's treated early at home, and we can prevent those costly hospitalizations, and eventually even mortality or that. So the story is I'm trying, I always say, I'm trying to solve a problem that killed my grandmother. So my abuela when she died, was because she wasn't very good at telling us, she was declining. And when we picked up the signs where we're obviously too late, she got into the hospital, and she died from complications from these an ICU admission. So the whole thing will have a goal that prevented

Nick Talamantes  4:07  

First, lo siento mucho por tu abuela. It's tragic to lose someone to something that could have been prevented. However, you're using that personal motivation to now offer hope to millions of other patients in the world, which is incredible. You mentioned if you catch it early, you can monitor and intervene. How early do you believe your technology will be able to to identify these signs of an exacerbation coming?

Maria Artunduaga  4:31  

Sure, so I've talked to a lot of people like over like 200 250 people in the respiratory space. So in order for us to actually be commercially adopted, just a matter of like 36 hour, it's like, you know, the cutoff. But the thing is, again, I go back to like, predicates like historical, historical precedent order companies, have demonstrated that by using digitizing biomarkers, you could even do it for only five to seven days prior to the exacerbation when it's diagnosed itself. So again, I mean, if we are able to 36 hours, that's more than enough for for actual companies or doctors or two, three, you know, buy it or start using it. So one of the main objectives wheeling the next 24 months is to actually run an exacerbation trial with a prototype that's robust enough on a platform that's robust enough and can catch the signals 24/7 for you know, several days, you know, and we still need to work a lot on the engineering part of it. And ideally, to see if we are able to see those signals that are different and how that turns into like, exacerbation how it's diagnosed, if doctors you know, realize as an actual true exacerbation on how we can prevent a hospitalization event. So that's a long, it's a long way. You know, everything I'm gonna go technology takes forever. But that's a goal. And yeah, I mean, it. I love the fact that it's challenging. But it's so important that we don't scare away from things that are difficult.

Nick Talamantes  6:20  

Absolutely. You have to address those challenges. Yes. So maybe speaking about those challenges. You mentioned that there are other competitors, other products out there today? What are sort of the limitations of those technologies? And how are you distinguishing yourself and improving upon imagine what they're doing already?

Maria Artunduaga  6:38  

Sure. So there are a few things here. So the status quo the standard of care to that you think that we are fighting against, or competing against, literally, our questioners. So questioners so for example, in COPD, you have something that's called a cat assessment, it's a list of 10 questions, and you score each question from zero to five. That's what killed my grandma, right? We have subjective data, questionnaires to diagnose exacerbations, that's how doctors how we figure out if you are exacerbating, and those questionnaires are missing 50% of the exacerbation, so half of exacerbations are never diagnosed. That's why it's so costly, and people are dying so much. And then there are a few other things. So then the patients realized, I want to know numbers, somehow I want to figure out they call it I want to know my numbers, they have pulse oximeters. So doctors, we don't really like pulse oximeters. But they don't tell us anything about the lung. It's a lagging indicator of a problem. And when you hit a number, which is 88, you know, this person has to come to the ER or we need to do something like 911 right away. So patients who post oximetry, and as I was saying, it doesn't tell you much about the lung, so doctors, we don't really do make a lot of clinical decisions based on that number. And then the gold standard, what we've been doing or using for 70 years is this pyrometry. What it is, is it's a device that you have, and you blow through it for several for 10 seconds, and you get a lot of numbers. The problem with us pyrometry is that because you have to do a forceful maneuver. Patients really dislike doing it because their lungs are not okay, their lung function is reduced to like 70 50%. So they are so scared of exacerbating symptoms, that they just don't do that. And doing the whole maneuver. It's it's hard to reproducibility it's a problem. So when you talk to pharmaceutical industry companies, right? They always told us, well, Maria thing is like, yeah, we have a spider metric for our clinical trials. But the repeater utility is so poor that we are looking for sensors. And right now, there are a lot of startups doing doing digital stethescope, so they just lease them. We are not a digital stethoscope. They just listen to the sounds that are coming from the lungs and in medicine. I mean, we don't really use a set of stethoscopes anymore. Not really, or at least doctors. We don't use them. Because we know that are better technologies, right? I mean, just listening to the sounds, even if it's digital. It's the diagnostic sensitivity is only 40%. So when you have COPD or asthma or many of these respiratory conditions, sometimes baseline wise, you have abnormal sounds, but it doesn't mean that we're exacerbating so it's like, yeah, I mean, yeah, it's abnormal. But then what happens is that you need to say or take this patient to do another test. So it doesn't really move that needle. We don't really use stethoscopes, I mean, I'm sorry. 

Nick Talamantes  9:53  

No, it makes sense, right? You can only tell so much from a sound. Exactly, exactly. So Have you guys are not just listening to sound, you're using artificial intelligence to decode what that information means? Are you training it internally based off of data that you're collecting? Or is there a larger data set? How are you making your AI?

Maria Artunduaga  10:17  

Let's talk about it. Yeah, sorry. I'm so geeky. I love it. And I'm not an expert on AI. But it's so interesting. So the main challenge that we have as a company and I sometimes say is, this looks more science for a company itself, at least not not right now. Nobody has ever tried acoustic resonance to figure out the lung ever before. So that's good and bad, good, because we had already secure six patents. And that's great, right? Because we have, you know, an area that we can protect. But since nobody has ever done it before, we need to build a database. So right now, we are testing a lot of people, six to eight patients every single week, in one of our sites in Florida, and we are starting to test probably 60 to 100 people more in California. So our goal is to have at least that, uh, from 150 to 200 by the end of this year. And even though it's not like 1000 people to trick us, you know, train models, you need 1000 5000 people. I mean, ideally, we have so much data per patient, we have two and a half hours of data from itration. And that's where it's, you know, exciting, but also is taking long, because crunching the numbers takes forever. But I mean, we think that by the end of December of this year, actually, we could probably have very meaningful results. And I'm excited.

Nick Talamantes  11:48  

I'm excited to hear more about it, too, when the results come in. Maybe shifting gears a little bit here. Do the acoustic biomarkers that you're collecting, do they are they the same for respite, chronic respiratory patients across the different disease states? So asthma, you're looking at the same biomarker biomarkers as you are in COPD?

Maria Artunduaga  12:09  

Yeah. That's so interesting, because it's, it's very pathophysiological the question. So we know by science that, I mean, it's so interesting, that our we call it phenotypes. So even amongst the same type of disease, there are small groups that behave differently. So I don't think I mean, we have already seen your data that COPD behave certain way and they like cluster here, asthmatics do something here on lung, oh my gosh, they're all over the place. So because lung COVID We don't know anything about lawn care, right? It's like new. But that opens up on our door for like research and development like anything drug development, therapeutics. I'm super excited, because even though like oh yeah, we don't know. In the next couple of years, the next decade, we can potentiate all things research with our technology. So right now what we're seeing is that COPD is behaving in a very particular way. So the Acoustic Resonance profiles are different from the others. So that's one thing that I can tell you today, for sure. That is that's what we are saying.

Nick Talamantes  13:27  

So then are you focusing right now only on COPD? Or are you working on multiple training your AI on multiple different disease states?

Maria Artunduaga  13:37  

I would love to say that we are doing a lot of things but you know, as a company, you need to be focused, right? I mean, if I had the money, I will do asthma and lung COVID on COPD just because of the need and medical hat on. But right now we are focusing on COPD because there aren't a lot of competitors doing COPD. They need clinical need and economic need. It's obviously higher. And I think I mean, I have my own personal bias. Right. My grant my abuela died from COPD, I really want to fix this problem. But I mean, the future definitely asthma lung COVID. We are, you know, contemplating doing contracts for grants, where as DOD, NIH, or a few other things, but again, yeah, you're right. These are focus.

Nick Talamantes  14:30  

That's good. I think, you know, speaking of money, are you guys currently raising money right now?

Maria Artunduaga  14:34  

No, really. I mean, I came because I am a first time founder and I've never raised before like for real going out going out fundraising. So I thought this was a really interesting conference that people a lot of my friends were coming so it was like Okay, let's go and like see people after the pandemic, like a opportunity to dress up but yeah, no, great. So I wanted to like test waters and myself on this stage. It was, I was so nervous, oh my goodness. But in a way I wanted to, like, see how I do it. I wanted to test myself as a CEO. Do people like it like you're able to see people like the energy do people like, I needed to validate the company? Right? So we are not fundraising. But I'm open to have conversations, and especially feedback. I need to learn a lot. And I know, I need to know, what is it that moves the needle for investors to say, I want to join this team, I want to help Maria, to, you know, solve this problem. And yeah, I mean, I'm here to learn,

Nick Talamantes  15:46  

It's great that you have such an open mind. You know, having a humble attitude, when you're speaking with investors, is really important to pull in the feedback you need to, you know, make your mission a reality. Yep. As a first time founder, do you have any advice for other first time founders that are out there?

Maria Artunduaga  16:04  

Oh, my goodness. So many things. Well, I think what you are saying, be open to feedback, or criticism, what are how are you going to frame it? Don't fall in love with the technology, right? You are going to be with for certain at least three four times, we were originally going to spin out technology from UC Berkeley, where everything is said when I was doing my master's there. So be open to like, you know, what's your actual mission? Do you really want to solve a problem. And if you are, you should be able to modify a lot of things, and especially to listen to a lot of feedback. And a lot of suggestions. At the end, you decide what we're going to do or not decide if that's a good advice, or no, I'm going to follow through. But especially that because people or at least in Silicon Valley, where I live, I've never experienced that type of community or environment where people like literally for free help you with advice. And I'm so privileged to be there and have access to that people because I'm from Colombia, right. So I'm, I'm from a developing country, I don't speak perfect English, I have an accent, right. And he was even worse before. And on again, how many people from a letter American country has the opportunity to be right at the mecca of everything technology, right. So I'm so thankful for the opportunity so that I, you know, I, I asked for a lot of questions, I ask for a little help, and I'm getting it for free. I love that. Right. So I would say that and asking for help people think that when you're a CEO, you're supposed to be perfect. You're supposed to get everything right. Investors are really looking for the people who have the right answers, but the right attitude to say, if a door closes, I'm gonna find a backdoor. Yeah. And I do that very often. But still, you know, I'm, I'm so different from the successful founder, you're the profile that they are looking for, like regular, usually a male, usually somebody who is yeah, wide, who speaks perfect English, I look so different. So historically, there has never been somebody like me who has taking a company IPO, and even less in medical technology. Right? So I'm a high risk founder, right, I'm a high I have a high risk profile. So try not to fight it, you know, do not fight it, you know, find a way to get it done. So that's why I do a lot of grants too. That's why I'm talking to a lot of funds that in a way have some sort of signal or story that in a way, I feel connected to like, you know, the challenges, you know, the personal stories, right, like I do a lot of things because of my family. Right? So that's how you do it. Yeah, just keep persevering. I mean, this is gonna take forever,

Nick Talamantes  19:16  

You know, the adversity will make the final reaching the finish line that much more rewarding. Maria, you are such a vibrant and driven person. I know that you're bound for great things, and I can't wait to see where you and your company go in the future. Thank you so much for joining me in the studio. 

Maria Artunduaga  19:37  

Yes. Thank you for inviting us so much.