Rhoda Au on digital biomarkers and Precision Brain health
EPISODE SUMMARY:
Harry speaks with Boston University’s Rhoda Au, who believes that algorithms parsing new kinds of digital data about individual patients could find warning signs of diseases like dementia while they’re still preventable—leading to a new era in which precision medicine is gradually replaced by “precision health.”
SHOW NOTES
As one of the researchers involved in the data anyalysi of the 70-year-long Framingham Heart Study, Rhoda Au is in a unique position to investigate whether changes in speech patterns in middle-aged people could prefigure the onset of Alzheimer’s disease later in life, and whether early detection might give patients more time to take preventative measures. She’s been part of the Framingham study since 1990, and she’s applying voice analysis software to 9,000 digital audio recordings of neuropsychological exams of Framingham patients to see whether there were telltale biomarkers in the speech of patients who went on to develop dementia.
Au is a professor of anatomy and neurobiology at Boston University, a professor of epidemiology at Boston University School of Public Health, a senior fellow at the Institute for Health Systems Innovation and Policy at BU’s Questrom School of Business, and the Framingham study’s director of neuropsychology.
Please rate and review The Harry Glorikian Show on Apple Podcasts! Here’s how to do that from an iPhone, iPad, or iPod touch:
1. Open the Podcasts app on your iPhone, iPad, or Mac.
2. Navigate to The Harry Glorikian Show podcast. You can find it by searching for it or selecting it from your library. Just note that you’ll have to go to the series page which shows all the episodes, not just the page for a single episode.
3. Scroll down to find the subhead titled “Ratings & Reviews.”
4. Under one of the highlighted reviews, select “Write a Review.”
5. Next, select a star rating at the top — you have the option of choosing between one and five stars.
6. Using the text box at the top, write a title for your review. Then, in the lower text box, write your review. Your review can be up to 300 words long.
7. Once you’ve finished, select “Send” or “Save” in the top-right corner.
8. If you’ve never left a podcast review before, enter a nickname. Your nickname will be displayed next to any reviews you leave from here on out.
9. After selecting a nickname, tap OK. Your review may not be immediately visible.
That’s it! Thanks so much.
Transcript
Harry Glorikian: Hello, I’m Harry Glorikian. And this is Moneyball medicine. The show where we meet executives, entrepreneurs, physicians, and scientists using the power of data to reinvent healthcare from machine learning to genomics, to personalize medicine. We look at the biggest trends in patient care and healthcare management.
And we talked to people behind the trends to find out where data is making the biggest difference.
My next guest has said that because of rapid advances in technology, it is now possible to collect digital metrics, to assess, monitor, and detect chronic disease indicators much earlier. Data science and artificial intelligence can drive the discovery of digital biomarker before the emergence of overt clinical symptoms, thereby transforming the current healthcare approach from one centered on precision medicine to a more comprehensive focus on precision health and by doing so enable the possibility of preventing disease altogether
Rhoda is professor of anatomy and neurobiology, neurology and epidemiology at Boston university schools of medicine and public health, and is a senior fellow at the Institute for health systems, innovation and policy at the Questrom school of business. She also serves as the director of neuropsychology at the Framingham heart study, where she has been involved in research relating to cognitive aging.
And preclinical and clinical dementia since 1990, she’s very interested in how big data analytics can better inform our understanding of disease, pathways and treatment. In addition to our work at Framingham heart study, Dr. Au is also currently focused on building multi-sector ecosystems to enable solutions for chronic disease.
Prevention generally an optimizing brain health specifically, and to move to the primary focus of health technologies from position 80 medicine to a broader emphasis on precision brain health.
Harry Glorikian: Rhoda, welcome to the show.
Rhoda Au:Thank you.
Harry Glorikian:Tell us a little bit about all this fascinating work that you do.
Rhoda Au: Uh, well, I can start out by maybe talking a little bit about the Framingham heart study.
Harry Glorikian: Sure, sure. Excellent
Rhoda Au: Uh, so this is a, uh, an epidemiologic study that actually started in 1948 and many people actually aren’t aware of the Framingham heart study, but here’s the impact that it’s had on virtually everybody’s life. Right? When you go into a doctor’s office and you get your blood pressure measured. That’s because of Framingham, because in 1948, we actually didn’t know the concept of high blood pressure.
We actually knew that people were dying of heart disease and stroke, but we didn’t know why. And so the NIH commissioned the Framingham heart study as a 20 year study to figure out. And in 1961, Framingham published the first paper to identify risk factors for heart disease and stroke. And those are many of the common risk factors that you now have measured when you go to your doctor’s office further in that 1961 publication, they also define the concept of risk factors. They actually coined that term and that actually opened up the entire field of preventive medicine. So that’s really the impact that Framingham has had, uh, from its very beginning.
Harry Glorikian: Interesting how you have to have a, almost a government study, a funded study to sort of pull that sort of magnitude off as opposed to a, uh, privately funded or a company doing it themselves.
Rhoda Au: Sure.I think at that time, you know, we have to give credit to our government leaders, uh, to be prescient enough to realize that it was going to take that long. And so they did commission it initially as a 20 year study. Now, interestingly, now, today it’s actually just finished celebrating its 70th year. And so, uh, and what’s happened is in following this, uh, original cohort of participants, uh, from 1948 until now, and then bringing in their children and their children’s spouses in 1971 and their grandchildren actually in 2002.
Um, and then also expanding to a couple of multi-ethnic uh, cohorts, uh, because at the time at the original recruitment, Framingham was largely a Caucasian population. And so they want to expand the diversity of that. But what’s happened is, is if you follow groups of people very closely over time, even though they started out as a study of heart disease and stroke, it’s actually morphed into a study of common chronic diseases.
Harry Glorikian: Right, right. Yeah. Yeah. Yeah. I can imagine because while you’re looking at these and I think the body has a limited number of biomarkers that it can show you that at least noninvasively that they have to, it has to mean other forms of disease that they’re all sort of pointing towards, it’s not just one per se.
Rhoda Au: Well, I think in a natural, you know, the point of an epidemiologic study is to follow the natural course within a community-based sample. And so this is just following the data, right? This is just following people as life unfolds itself over decades of time. And so even though you have the focus of heart disease and stroke, as you say, you know, people develop many other conditions.
And so we’ve been able to track those as well. And that actually sort of led to my, the work that I do, which is really focused on the brain because it turns out what’s bad for your heart. It’s also bad for you brain
Harry Glorikian: Yes, no. Well, and I’m glad that, that, uh, give a shout out to Steve Tucker who was kind enough to, uh, introduce you to me, but, you know, he he’s focused when, when I was talking to him, he was very much talking about, uh, your focus on using voice as a digital biomarker.
So can you tell us about your work in that area and where, where you see that going based on the work you’ve done so far?
Rhoda Au: Sure. So, uh, so when I entered the field of neuropsychology, I was really concerned about the fact that the cognitive tools that we have today are relatively (inaudible). Uh, relative to, you know, representing the full, uh, comprehensiveness and complexity.
Harry Glorikian: Right.
Rhoda Au: Over cognitive capabilities and also every neuropsychological test, even if we define it as a domain specific test, in fact, it’s multi-domain in its execution.
And so I was always very concerned about to what extent were we really accurately, uh, picking up, you know, this cognitive information in a way that really represents someone’s cognitive status. And so what happens is, is when you give, uh, neuropsych testing to, uh, participants. You know, there’s the response that you want, which is quote unquote the correct response that you’re looking for, but it turns out that they give you lots of responses.
Harry Glorikian: Right.
Rhoda Au: And what happens is in the traditional way of scoring these neuropsychological tests, you traditionally ignore those extraneous or error responses. And there’s a way of [looking], thinking about neuropsychology, interpreting people’s responses as called the Boston process approach. It was really championed by Edith Kaplan, who was at the Boston VA at the time.
And I really sort of embraced Uh, what Edith had taught us about not thinking about what was the person’s final response, but what was the pathway to that response? And so, consequently, I wanted to capture all those extraneous responses, but when you’re actually testing someone, And you’re, you know, reading them questions and they’re responding. You actually can’t write everything that they say down.
Harry Glorikian: No, that’s impossible.
Rhoda Au: Right? So, so if you really want to capture everything, they’re saying you have to think about recording it. And because this is Framingham and it’s a longitudinal study that goes on forever. You worry about saving everything.
Harry Glorikian: Right.
Rhoda Au: And so what happened was, is that I realized I needed to record these people’s responses. In order to catch them all. But I also didn’t want to clutter us with like CDs and tapes, et cetera. And so at that time, the possibility of saving files digitally was an option. So we started to voice record people’s responses.
Harry Glorikian: Those must have been huge files though when I think about-
Rhoda Au: Uh, yes, they were. And that’s part of why, because it would co-, you know, it would be lots and lots of CDs.
Harry Glorikian: Right. Right
Rhoda Au: And so how am I going to store that volume of data easily?
Harry Glorikian: Right.
Rhoda Au: And so we decided to record it digitally. And, uh, so we started that in 2005. And then when Siri came out, I realized that those voice recordings that I had been doing, basically for quality control purposes, right. To capture and be able to write down everything people were saying were data themselves.
Harry Glorikian: Right.
Rhoda Au: And I realized voice recognition, voice analysis software had become very sophisticated. And that’s where the concept that there’s actually the possibility of digital voice
Biomarkers embedded in these recordings. And then I spent the next few years trying to convince people that in fact, this was really, uh, interesting data.
Harry Glorikian: That’s interesting that it took you so long to convince people of that because voice prints, uh, looking at modulation of voice emotion, all of that is embedded in the sound that comes out of our mouths and what we say and how we say it.
And the delay, all that is embedded in it that piece of data. So you would think people should have caught on to that pretty- a lot faster than maybe they did.
Rhoda Au: Yeah. I think that it’s always the case in science that we have our methods that we know. And when we start to try to embark and methods that we don’t, uh, that is a little bit more difficult to embrace.
So I would say my history is always that it’s been hard for me to get people to embrace ideas that I have. Um, but, uh, But I’m persistent. And, um, I was fortunate to connect with, um, uh, Deb Kilpatrick from Evidation health,
Harry Glorikian: Yep
Rhoda Au: At a meeting, where she listened to me talk about these voice recordings. Cause I was going around and telling anybody about these voice recordings .And she actually was the one who decided that, wow, this is maybe something worth pursuing. And we brought in Jim Glass from MIT. Who had been doing, uh, this kind of digital voice, uh, um, work for many, many years. In fact, I think he has worked with some of the early foundations that led to Siri
Harry Glorikian:OK
Rhoda Au: And we got a DARPA grant together and were able to start exploring, uh, these voice recordings.
We actually today have accumulated almost 9,000 of these. Over this period of time of which some of the people of which we recorded from initially have gone on to become demented. And so we were able to then look to see, can we, on the basis of these voice analysis, start to separate out those people who, who are showing signs of cognitive impairment versus those who haven’t.
Harry Glorikian: And so where are you in that analysis? Do you, do you feel you have a good enough set of traces that you can use that can match up to a disease state.
Rhoda Au: I think we have it embedded in there. This is a lot of work actually-
Harry Glorikian:Yes
Rhoda Au: To do this kind of-, so we had done it on a subset of recordings or actually about 200 of these recordings. And as I said, we have about 9,000 of them.
Harry Glorikian:Yep.
Rhoda Au:So, uh, so we’re in the process of sort of trying to find additional support, additional interest and move this forward and actually. Uh, you know, now the concept of digital biomarkers, which when I brought it up, you know, seven, eight years ago, uh, people didn’t think was, were two words you could put together.
Harry Glorikian:Yes
Rhoda Au: Uh, now has some traction and even the FDA is talking about the concept of digital biomarkers,
Harry Glorikian: Yeah
Rhoda Au: Which is opened up, I think, opportunity for those of us who are lucky enough. To have, uh, in advance acquired a number of these digital recordings. Um, you know, we’re sort of now in a, in a good position to be able to look and really start to test and validate these concepts of digital biomarkers.
Harry Glorikian: Well, there’s a, there’s a, there’s a few companies that have, that are out there that are utilizing voice for, um, emotional state, um, and hopefully disease prediction or, uh, being able to see something, you know, way before, or it’s in it’s earlier stages. I don’t know which one of them is going to make it necessarily all the way to the finish line, but it is a very exciting sort of area.
And I almost think though, it’s not going to be one digital biomarker, but a combination of digital biomarkers that come together that might give you a better picture of what’s happening with that patient.
Rhoda Au: Yeah. I think that there’s a couple of few things. So one, uh, at least within, let’s say the realm of Alzheimer’s disease where we now don’t have any effective cures today.
So we have to start thinking about, um, this idea that it’s an insidious onset disease. So it’s not like the day before, the week before the year, before, even five years before, from the time of diagnosis that a person was likely to have been normal. In fact, they’ve been probably showing some indications of, of disease onset for a while.
Harry Glorikian: Right.
Rhoda Au: And so. Uh, so the question becomes whether these digital technologies allow us to capture, uh, changes in what I like to think about is if we are collecting data on a continuous basis and monitoring it, And if we can start to detect when that first negative trajectory of change emerges and potentially well within when what we would still call normal, if we could then intervene at that point, maybe instead of delaying the onset of the disease, uh, maybe we can change the trajectory and prevent it all together.
Harry Glorikian: If we understand, I can, I can see on the diagnosis side, it’s trying to understand what drives the mechanism of the disease, which may result in the right therapy.
Rhoda Au: Well, I think what we have to do is stop and think about our presumption of, is it a disease? So in the case of Alzheimer’s disease, we know that there’s a lot of heterogeneity and its expression over, the course of the disease from beginning to end.
So for me, I’ve always thought about the fact that it seems like it’s multiple diseases, not one disease
Harry Glorikian: Yes
Rhoda Au: And so what we need to first do is we need to figure out how do we embrace the entire complexity of the disease and really figure out how to characterize it
Harry Glorikian:Right
Rhoda Au: And then the other thing, I think that when it comes to continuous monitoring of an insidious onset process, at what point do we call detection of the symptoms diagnostics
versus when do they become prognostic?
Harry Glorikian: Right.
Rhoda Au: You know, where we cut that is actually a little bit arbitrary. And so when I think about digital biomarkers, I think about the fact that sure we have the opportunity to think about diagnostic, but again, if we can actually catch it early enough, are we really talking about now prognostic?
Harry Glorikian: Right, well, I think about companies like Cardiogram that have done those large studies with UCSF. And it’s just, you know, my heartbeat off my watch. And, you know, they’re able to give you an indication of, uh, you’re pre-diabetic you have sleep apnea. You might have an arrhythmia, it’s not a diagnostic, but it’s go have this validated by, you know, the system.
Um, you know, that’s one single biomarker. Could you take that biomarker along with voice, along with handwriting or other thing that would affect, uh, the neurology of the patient, or give you an indication of their neurological state of the patient, and then be able to do a, get a better understanding of the subcategories of that disease.
I, I, whenever I hear of dementia and Alzheimer’s, I always think those are super words like when we used to say breast cancer and now there’s well, which breast cancer is it? And I feel like the neurological world is sort of going along the same path.
Rhoda Au: Yes, I agree. I actually like to use cancer as a model of yeah.
Before we used to wait, if you had a diagnosis of cancer, it was a terminal event and, um, and people just went home to die and that’s, uh, but today we think about what type of cancer, what stage of cancer, you know, just for leukemia alone, there’s over 70 different treatments. And I really think that that’s where we need to go get to get to where we need to go for finding effective treatments for Alzheimer’s diseases.
Rhoda Au: Right. So I do think that precision medicine is actually the concept of the right treatment for the right person at the right time.
Harry Glorikian: Right.
Rhoda Au:What I like to think about more broadly is precision brain health. Which is really about the right solution for the right person at the right time.
So rather than focus on the disease model, which is really the medically-oriented model, because you’re still talking about symptoms of a disease. What I like to think about instead of, instead of focusing on Alzheimer’s disease, neurodegeneration, et cetera, why don’t we think about maximizing brain health? Why don’t we think about cause I’m a neuropsychologist,
Harry Glorikian: Oh yeah,
Rhoda Au:So I’m interested in peoples. Uh, cognitive status, wherever they are on that continuum. Um, and I think that if we start to monitor on a continuous basis and not think in terms of disease, but in think about, we want to optimize your brain health, wherever you are on that continuum, normal to disease.
And I think that that’s sort of a different shift in thinking about how we want to approach and how, where I think technology has a really transformative opportunity that goes far beyond what precision medicine is. Because then we start to think about, uh, people, when you start to think about optimizing brain health, you don’t think about it as something you do later in life.
You think about something that you can do across your entire life. So I think about the idea of brain health as something that’s a womb to tomb experience.
Harry Glorikian: Right. Well, -knows. I could use as much optimization as I could get. Um, but what are the sort of, uh, technologies that you’re utilizing or that you’re seeing people utilize from.
Uh, either technology or sensor or combination thereof, uh, in, in, in this area that you see will be transformative.
Rhoda Au: Yeah. So I actually don’t think it’s any one technology, I think, as you said, I think it’s going to be a combination of technology. I think digital voice, is one piece of that, what I like about digital voice though, is how ubiquitous it is.
Um, because one of the things that we have to appreciate is that the study, um, Measuresment Assessment of cognition is something that’s very mature within the U.S. It’s very mature in Europe, but it’s not elsewhere in the world.
Harry Glorikian: Right.
Rhoda Au: So for instance, in China, where I do a fair amount of work, uh, they have the most rapidly aging population in the world and they have no PhD programs in neuropsychology.
So they don’t actually have. The expertise internally to service their population and cognition tends to be a very culturally specific, uh, kind of, of test, right? And so are the tests tend to be very culturally specific. And so for me, when you think about digital voice, you think about the fact that most people speak so independent of their education. Of their race of their language, of where they live. You know, so if we think about trying to reach all parts of the world, digital voice is a very powerful tool. And when I think about putting together a brain health monitoring platform of multiple technologies, I think about that, I think about how do I go about reaching people wherever they may be.
And regardless of whether the expertise is present in that country. And so that’s what we’ve been developing is it turns out that the smartphone is a really, really powerful tool
Harry Glorikian: Very
Rhoda Au: In which to think about deploying that. So we, so I’ve developed actually a brain health monitoring system, and I focus on some fundamental behaviors, physical activity.
Sleep, medication use, social interaction. Right. Um, and, uh, I’m trying to, and, and in this case cognition.
Harry Glorikian: Right.
Rhoda Au: Uh, which includes sort of traditional cognition, as well as things like digital voice. Which can allow us to get away from, uh, traditional testing of cognition to inferring cognition.
Harry Glorikian: Understood.
Rhoda Au: Okay. And so, uh, so we have been developing sort of an integrated plug and play system with the idea that it could be deployed. And here’s what I’m, why am I interested in those particular activities? Because they’re common to everybody and at the end of the day, Everything we do, we do through our brain. So when we are going through our natural environment, you know, engaging with people, taking our medications, sleeping, et cetera, we’re actually always reflecting our cognitive status.
And so my goal is to get away from testing cognition in an artificial way to inferring cognition through people’s natural behaviors.
Harry Glorikian: Yep.
Rhoda Au: And that’s really how our brain health monitoring platform is really set up
Harry Glorikian: Well It’s usually in the field where a loved one notices that something is up with.
Rhoda Au: Correct.
Harry Glorikian:Right. So they’re inferring that something doesn’t look right. And so if you could have technology sort of quietly monitoring that, um, that would be interesting. I think a voice has almost Active monitoring because someone’s got to engage with the device, whereas, you know, my apple watch or something like that is it’s just passively monitoring me as I’m going.
So I always think about people, uh, uh, following through on what we would like them to do, making sure that they do those things.
Rhoda Au: Yep. And so what I would say is where’s technology today, most health technologies. Are built with the patient in mind. And so they require a lot of active engagement, they kind of called high friction. Right. Yeah. And what we really need to get into is what you were just alluding to, which is much more ambient technologies that are picking it up passively as people engage in their environment, but don’t actually require them to do things in order to, uh, collect the information.
Harry Glorikian: But, how do we bill for that? Can you imagine, how would the healthcare system business model be effected by finding people before they get sick.
Rhoda Au: So that’s because you’re presuming a medical model, right? The medical business model again, assumes the patient. And so in order for the medical community to. Embrace it, as you say, there has to be a billable code.
Harry Glorikian: Right?
Rhoda Au:So what I believe is that we’re actually going to have to create a whole new business ecosystem. That’s going to be complementary to the medical one. And so I always say this when companies like Microsoft, apple, Google, IBM, and they’re all working on it-
Harry Glorikian: Right. Right
Rhoda Au: As well as tens of thousands of different technology companies, when they rise up and become equal players in the healthcare space.
We actually will have started to solve this problem. And here’s the thing that I want to remind everybody. 86% of diseases in the US are chronic diseases. And that means, that means that the majority of them are insidious and onset,
Harry Glorikian: Right?
Rhoda Au: And the reason we have high healthcare costs and low healthcare qualities is because with the medical model, we wait till your symptoms had hit a certain threshold.
And then we’re really good at treating it. For a really long time.
Harry Glorikian:Right. Right.
Rhoda Au: And that’s why we can’t reverse that trend
Harry Glorikian: Early detection is not, um, smiled upon by some people. Right. Because it sort of kills them. Their revenue.
Rhoda Au: Right. And I think that we have to accept, we’re always going to need doctors.
We’re always going to have, patients, there are always going to be people who are, who no matter what they do, they’re going to get sick. Right. And we need that. We still need that. But at the same time, we need to recognize that if we really want to if we really, really, really are serious, about controlling our health care costs and really serious about increasing our health quality, which only rakes 37th
Harry Glorikian: Right
Rhoda Au: In the world, even though we’re the highest per capita spending, we really have to think about a different approach and that’s a different mindset.
It’s a shift away from the medical model and it’s actually much more one of a public health model.
Harry Glorikian: Well, I know, I think I’ve always thought about it. Before I had put together the book I wrote, it was always thinking, well, how do we bring six Sigma to healthcare? First of all, just, just procedures that we’re doing already.
How do we capture the data? Do the analytics go back and say, and, and optimize that. And I know we call it best practice, but that’s not it. You need a real rigorous analytical model where you’re constantly fixing what you’re doing. And The low-performing items. You, you, you can’t keep them cause it won’t get you to your six Sigma.
Um, and then you slowly move to preventative care. Right. But, but there’s a lot of, I think that there’s an incredible level of low-hanging fruit from a cost savings perspective. If, and I bring this example up, but a Geisinger has a money-back guarantee on surgery. Right? So what if every institution in the United States.
Sort of adopted that model and I’m not saying they need to create a new one, just adopt the one that’s already being used. What would that do to healthcare costs and then technologies like you’re talking about where we can, you know, detect something much earlier, like this Cardiogram, uh, app that I talked about.
If you could identify everybody that has sleep apnea earlier wow. And treat them that would have. That would have significant impact on a number of chronic disease areas.
Rhoda Au: Right. Or if you could detect that people are starting to have sleep disturbances before they even reached that state. Then you can intervene even earlier with probably interventions that are even less costly.
Harry Glorikian: Right.
Rhoda Au: Uh, and then again, it’s about changing that trajectory altogether,
Harry Glorikian: Right?
Rhoda Au: So it’s, it’s, it’s again, it’s, it’s a big mindset shift for people to think about the health part of healthcare, because we tend to think about the care part and, uh, and it’s really about how technology, and I’m not a technologist.
I may like, I make my smartphone dumb. So, so, so for me, you know, technology is really the tool. And then, and it really allows us the opportunity now to monitor our own behaviors in a way to allow us to intervene at a time where we can maybe then keep ourselves.
Harry Glorikian: Right. So where do you see the work you’re doing at?
What stage do you think its at? Where does it need to go next? And then what do you see the impact of it? Um, and again, Maybe it’s not just voice, but all the different methodologies that you’re using and the impact that it’s gonna to have.
Rhoda Au: Sure. So I think there’s a couple of things that need to be done.
I think we’re a long way from where I’m talking about, but I, uh, I think that even though a long way, uh, technology has a way of compressing the time.
Harry Glorikian: Right
Rhoda Au: In which we can do it. So I’m a big proponent of open science data sharing because this is our path. This is what I think is the path to digital biomarkers.
I think that we can collect all this data. There’s a lot. We won’t understand there’s a lot. We don’t know. And quite frankly, I’m not going to be the person to figure that out. I’m only a neuropsychologist. So what I need to do instead is aggregate this data in a responsible way, protect people’s confidential information, of course, and then I need to release it and I knew they would release it to the data scientist and Artificial Intelligence community, wherever they may sit, whether it’s in academics for profit, et cetera. Again, precision brain health is going to be about the right solution for the right person the right time. So we’re talking about thousands and thousands and thousands of solutions. And what I imagine is things is that we can open up this database and issue channel.
You know, we have really smart people all around the world and if we can aggregate and give them the right data to work with, they’re going to go out and they’re going to discover all these digital biomarkers. And that’s why I’m a huge proponent of making sure that as we collect the data, we do it in a responsible way, but we also do it in a way with the intent of sharing it.
And, um, the other piece I would add to this and particularly in the, with respect to Alzheimer’s disease. We actually don’t even know completely how complex of a disease it is. So what we need to do in embracing that complexity is I think we need to go beyond the study of Alzheimer’s to get to the answers of Alzheimer’s.
So what do I mean by that? So for instance, at the Framingham heart study, I mentioned we study all these common chronic diseases. So we have people that we know develop cancer. And we looked at the relationship between cancer in the brain and we see a signal it’s there, but you know what? It’s a faint signal and why? Because we actually don’t do a very good job of characterizing cancer,
Harry Glorikian: Right
Rhoda Au: Cause we’re not a cancer study, but you know what? The national cancer Institute, they do really good jobs of studying cancer. If I can take that brain health monitoring platform apply it to these well-characterized cohorts of cancer, then suddenly I’m in a position to discover what’s that relationship between cancer and the brain.
Harry Glorikian: Right.
Rhoda Au: Okay. So that’s why I think that we need to think more creatively about how we’re approaching these complex diseases and, and not think about it sort of. Uh, uh, a certain pathway, but realizing that there are multiple pathways and how are we going to find what we don’t know? Because most of us in the field, we study the same kinds of things,
Harry Glorikian: Right. Healthcare Expert Investor for Digital Health & AI Machine Learning
Transformation of the Healthcare Industry Precision Medicine How Machine Learning is Changing Rare Disease Diagnosis
Rhoda Au:Because we think these are the things that are related to the disease. But what about all the things that are related to the disease that we don’t know?
Harry Glorikian: Well, it’s interesting you say that I always reflect back on, uh, an interview that I did with Uh, Joel Dudley from Mount Sinai where he, threw all the data in there and he pointed in a particular direction that how’s that possible that it, you know, and so I think some of diseases are so multifactorial and things don’t always start where you think they should start, that you almost need
Artificial intelligence, machine learning analytics, a combination of them all to sort of help you elucidate the path of how that disease got started. If you’re going to prevent it, or you’re going to change the path of use as you say.
Rhoda Au: Yeah. I mean, again, if we look at Framingham heart study, right study of heart disease and stroke, it’s being repurposed.
I think it’s, we have the richest cognitive aging and dementia dataset in the world. Even though we didn’t start out as a study of Alzheimer’s disease. And we now know that cardiovascular risk factors seem to have an impact, seem to have, you know, increased people’s risk for the disease. And Framingham is the best study to look at that.
Harry Glorikian: It’s interesting. You know, There was a, a, a commission study to be able to do this. And now I think in some ways, technology allows you to pull data or get data from patients or people without necessarily having to do a study like this, because they’re willing to share data with you. They’re willing to open an API to the, you know, uh, heart data that comes off your apple watch or your, or their weight scale or, or things of that nature and not dissimilar to what Evidation is sort of has a pulse on.
Um, so I always think that you can get more data faster because now the whole world is hooked up.
Rhoda Au: Yes. And so I’ve teamed up with another group of philanthropic group and we’re trying to create a digital trust with that idea. So one of the things I think we need to get away from us researchers is we need to stop thinking that we own the data.
We don’t own the data. The only people who own the data are people who give their. So people own their own data. And then they give us the privilege as researchers to use that data and to use it responsibly and to be good stewards of it. And I think they are in this digital trust. What we’re hoping to do is allow people to share their data for a common good in order to unleash the possibility of these discovery of these new digital biomarkers, uh, by people who, as you say, you know, if you put it in the hands of people who don’t know one way or another and, um, and just bring in their data science, you know, capabilities, um, and just, uh, and have at it, they’re going to discover things that those of us in the field never thought about.
Harry Glorikian: Right. I think we, I think those of us that are in the field need to validate those things.
Rhoda Au: Oh no absolutely
Harry Glorikian: Right. Um, but they will find things that we may not have identified.
Rhoda Au: They’re going to find different ways to do it.
Harry Glorikian:Right.
Rhoda Au: Cause we need new approaches.
Harry Glorikian:Right,
Rhoda Au: And we need much more again, we gotta start embracing the complexity of the disease.
Rhoda Au: So a lot of traditional research tends to be very reductionist
Harry Glorikian: Yes
Rhoda Au: In its approach. And I think that we need to be much more embracive and comprehensive. And so I think that this is also going to be a very, very big paradigm shifting, um, uh, period in which there’s what we did do versus what we need to do.
Harry Glorikian: Well and I think people now need to go beyond their area of study. I mean, I can’t read enough about the different techniques in machine learning or, uh, you know, different, uh, uh, AI methodologies and neural networks. And should we be using, you know, old fashioned rules-based, uh, algorithms and then you the combination thereof, right.
Which will get you to an answer. Cause I don’t, I don’t think the brain actually does one thing. The way that we design these systems, it’s a. It’s a multitude of different parts of the brain that come together that help you do what you do on a daily basis
Rhoda Au: Yeah. And I would say as a researcher, I often find myself in a room where it’s like, wow, I should not be here because I definitely don’t have any of the competencies that we’re talking about that are needed.
But what I do see is a vision that we need to bring all these different, uh, people together, uh, to try to solve the same problem. And so for me, it’s always, I think earlier you talked about I’m very. Interested in building these multi-sector ecosystems, because I understand its not one discipline or one group of people that are going to solve this.
And so I have learned to be very comfortable being a room where I know for a fact that I know the least amount of what we actually should do. Um, but I understand that the, these are all the people that need to be in this room.
Harry Glorikian: Well, it’s like putting a man on the moon. It wasn’t one discipline. It was a multidisciplinary effort.
So. On the lines of say voice or other technologies that you’re working on. How close do you think we are to seeing, I don’t want to say implementation cause that, that, that can mean something grandiose, but just some areas where you really see it moving the needle.
Rhoda Au: Oh I think we can move the needle within the next year or so.
I think that, you know, personally, I think we sit on. We- as in at Framingham, I think we sit on some data. Uh, we need more hands on deck to help us understand that data. But again, I think that, uh, technology has a way of compressing the timeline
Harry Glorikian:Right
Rhoda Au: Of what we’re trying to do. And so I actually am very optimistic that we’re going to get to.
At least some solutions, not all solutions, but at least the beginnings of some of these solutions, uh, much earlier, uh, than we might think right now
Harry Glorikian: I hope so because I’m getting older and I keep waiting for all these solutions to help me stay healthier longer.
Rhoda Au: Yeah I feel some urgency. I mean, there are people who are suffering from this disease right now,
Harry Glorikian: Right
Rhoda Au: Yeah and there are people who are right around the corner who are about to be diagnosed.
And so I feel, I’m very thankful that technology does make us move much faster than we’re used to, uh, because we need to get there in a hurry
Harry Glorikian:We might even be able to use this recording in the future to tell if there’s a, you know, early disease onset. But if you think about it everything’s being recorded, right.
So you could just unleash. The app on any recording?
Rhoda Au: Yes. Yep. And I will confess, I do worry about that the, the, the beginnings may have already started to sink in.
Harry Glorikian: Yes, yes, absolutely. Absolutely. Well, there’s a few people. I think we should be testing. No, we’re not testing. So. Well, this was great. Uh, thank you so much.
I really, I hope that your research goes wonderfully well and, and all the groups you’re working with help drive this whole process forward.
Rhoda Au:Um, well, I appreciate you giving me the chance to talk about it. And, um, like I say, I, I build these ecosystems and multi-sector partnerships and I’m always looking for more.
So there are people who are interested, they should let me know. Okay.
Harry Glorikian: That sounds great. Thank you.
Rhoda Au: Thanks.
Harry Glorikian: And that’s it for this episode. If you enjoyed Moneyball medicine, please head over to iTunes, to subscribe, rate, and leave a review. It is greatly appreciated. Hope you join us next time until then farewell.