Collaborative Endeavors

"Fine-tuning" innovations to predict manic episodes in bipolar disorder

Episode Summary

Dr. Alex Leow talks about her work in psychiatry and how her passion for playing the piano led to a career in neuroscience and spawned the idea to use smartphone keyboards to predict manic episodes in bipolar disorder.

Episode Notes


Alex Leow, MD, PhD
Departments of Psychiatry and Bioengineering
University of Illinois Chicago
Twitter: @alexfeuillet

BiAffect Project:

Steve Jobs’ Stanford Commencement Speech:


If you would like to see your interdisciplinary team featured on the podcast, reach out to me at

Interested in volunteering to participate in health research? Today’s researchers want to make sure that treatments and cures are designed for everyone’s unique needs. Are you ready to make a difference? Learn more at

The University of Illinois Chicago Center for Clinical and Translational Science is supported by the National Center for Advancing Translational Sciences, National Institutes of Health, through Grant UL1TR002003. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Episode Transcription


To the best of my knowledge, our BiAffect project is really the first large-scale research effort to look into whether smartphone typing would provide useful information about a person's brain health and mental health. And I want to just really thank our citizens scientists for being part of this effort. 

0:30  [Voiceover]

Welcome to Collaborative Endeavors, a podcast about how scientists from different areas of research come together to tackle big health challenges, leading to better therapies and healthier communities.

In our pilot episode, Dr. Olu Ajilore talked about his work in the area of depression and geriatric psychiatry. And as an aside, if you haven’t listened to this episode, I recommend you check it out to get the full picture of what this interdisciplinary team can do. Anyway- during our conversation, Dr. Ajilore praised his frequent collaborator, Dr. Alex Leow, and in the spirit of partnership, we thought it was time to hear her story. 

In this episode, Dr. Leow will talk about her work in psychiatry and how her passion for playing the piano- which you will hear through the episode- led to a career in neuroscience -and spawned the idea to use smartphone keyboards to predict manic episodes in bipolar disorder.

1:25 LEOW

My name is Alex Leow. I am currently a professor in Psychiatry and Biomedical Engineering. I have a hybrid appointment in the College of Medicine, as well as in Engineering. I am clinically trained as an adult psychiatrist. I’m an MD/PhD. My PhD training was in Applied Mathematics. And so, most of my research up until this point has been really thinking about how we can use computational models to better understand brain health.

When I was growing up, I actually wanted to be a professional pianist- classical pianist- or a mathematician. I was really fascinated by how we could potentially use mathematics to understand the world. But as you may know with Asian families, my parents really wanted me and my brother to be doctors. So, I eventually listened to them. And when I was in medical school, I really struggled to reconcile my innate interest in mathematics. I've also always been interested in reading books about general topics in neuroscience. When I was growing up, there was a really famous neurologist, Oliver Sacks. I read many of his books about fascinating presentation of neurological conditions in many of his patients. So, I thought about potentially using mathematics to understand not only how the brain functions, but what happens when the brain is not functioning.

At that time, I worked really hard to get into this kind of research because multidisciplinary research really only became more mainstream over the last ten or fifteen years. I struggled initially trying to combine medicine and mathematics so to speak. But I guess I was really fortunate to come across different mentors and eventually I was able to carve out this really unique and interdisciplinary space and the overall trajectory of where I ended up today. 

4:30 VO: 

To fully appreciate the “unique interdisciplinary space” in which Dr. Leow works, you need to take a step inside her mind. She painted a picture of the unconventional manner in which she views brain functionality and how her love for piano led her to examine ways to make it more resilient to aging.  

4:50     LEOW:

There are quite a few ways of conceptualizing this relationship with playing music- especially an instrument- and neuroscience and mathematics. Anecdotally it has been observed many, many times that a lot of musicians live a very long life, and they are still able to perform at a very high-level well into their eighties and nineties. So, there is something interesting about playing instrument. Once you master an instrument, it’s like supercharging a person's brain or configuring the brain in such a way that the brain is more resilient to aging. So that has always been a fascinating idea for me, again, because of my personal interest in playing the piano and also me as a neuroscientist. And again, drawing from my personal experience, when I'm playing the piano what specific aspect of playing the piano can translate to this resilience idea.

One thing I’ve noticed again and again, is this training in coupling two different types of memory systems. There’s what's called implicit vs. explicit memory. Implicit memory is things that we do but we can't necessarily describe what we mean. For example, brushing our teeth, how to use toothpaste. That is implicit memory. We’ve done it so many times since childhood, we can brush our teeth. But if you really want to describe what is being done it becomes hard to describe. And this is the exact opposite for explicit memory. Explicit memory would be something along the lines of, say, if I were to ask you your date of birth. You can declare your date of birth. I ask you in which city were you born? You can specifically name the city. What is interesting about playing the piano, at least for me, is this constant effort to merge these two types of fundamentally polar opposite memory functions

So the implicit part of playing the piano is, of course, muscle memory. I have to learn how to play these notes. But then there is a specific part- like if you were to ask me, “What is the first note of this music?” I should be able to tell you. I can say the first note is A for the right hand and, after that, followed by a cord with the left hand playing these notes and right hand playing these notes for this amount of beats and then what’s after that. So how do we actually combine those two types of memory into a cohesive and hopefully musical way? And that's fascinating, right? This constant training on how to really efficiently combine and learn a piece by leveraging these two types of memory and being able to learn it in a way that you could play in front of an audience under stress. It’s fascinating what the brain can do. And that kind of training can potentially provide a way for the brain become more resilient to aging. 

8:50     VO:

Fitness tracking devices, like FitBit and Apple watch, first hit the market around 2015. As a practicing psychiatrist and neuroscientist, Dr. Leow was fascinated by this new technology and began to ask herself- if we can make data trackers for the human body, why not the human brain?

9:10     LEOW:

And this is where my personal interest in playing a piano comes into play - no pun

intended. So as an amateur pianist, I am always noticing that how well I can play the piano is linked to my concentration and my mood for that day. So one day as I was playing the piano- and again, this is 5 or 6 years ago-  I soon realized that we should, of course, look to the keyboard of our smartphone to give you more context what I mean by that, imagine as I'm talking to you right now, I also start texting someone at the same time on the side. Now because my attention is divided, I will be making a lot of typos and my typing rhythm is going to be erratic or very variable. 

I joined UIC in 2014, and when I first joined- because of my dual appointment in medicine and engineering- the dean of engineering, Dr. Pete Nelson, invited me to his office on East Campus to have coffee. And then we started chatting and he asked me what my research interests are and since clinically I am a bipolar clinician- bipolar disorder, known as manic depression- we started talking about bipolar disorder. It turns out the dean actually has a personal interest in bipolar disorder and he said he has always wanted to do something in this area. And so when we were initially designing the app, again, coming back to this keyboard idea, if you are familiar with bipolar, when the person is manic they tend to be speaking really, really fast. And so again another intuition why we really think that the keyboard would work in this population is if they are talking really fast, then they should be typing really fast when they're manic. And so when we were initially designing this app the target population was bipolar disorder, people with this condition. And so what I did was we took the first two letters of the condition under the official DSM terminology, which is bipolar affective disorder, so it became BiAffect because of that.

We wanted to get as much data in a naturalistic environment as we can. And so once you download this keyboard, you can set it to replace your standard smartphone keyboard. And then every time you are texting someone or write an email or post on social media, the timing of your key press will be saved as well as things such as how often you use the backspace, how often you have a typo. Then we essentially try to build mathematical models to unobtrusively and passively infer a person's brain function. 

12:45   VO:

According to the National Institute of Mental Health, bipolar disorder affects approximately 5.7 million adult Americans, or about 2.6% of the U.S. population age 18 and older every year. It presents equally in men and women and is found in all ages, races, ethnic groups and social classes. Notably, bipolar symptoms can come and go- much like symptoms of depression and anxiety- though the disorder usually requires a lifetime treatment plan. 

Given the wide range of people affected by this condition, coupled with the fact that symptoms may fluctuate leading some to abandon care plans, one would think that research funding in this area would be plentiful. But, in fact, Leow’s team struggled to secure funding in the beginning. So they sought yet another innovative solution to their problem.

13:35   LEOW:

Initially we were really struggling finding research money to do this kind of creative, cutting-edge research. Intuitively it makes a lot of sense, but because of the very, very multidisciplinary nature of what we were trying to do that's so different than traditional models of research we were not able to convince any conventional funding agencies to give us the benefit of the doubt for us to try this idea out. I was about to give up until we were so lucky to come across this “Mood Challenge” that was co-sponsored by the Robert Wood Johnson Foundation.

And this is where the idea of open science really comes into play by leveraging technology, smartphones and connected devices. We realized that we can scale very easily if we can put this research app on the app store for everyone to use and they can directly download the app, learn about the study, and consent for the study anywhere in the US in about 15 minutes, then we really lower the barrier for anyone to become a participant of the study. And so we used the seed money to make this publicly available app. We put it on the Apple App Store and the rest is history. So that's sort of the genesis of our open science study. 

One fundamental principle of the open science model that is different than more traditional models of research. And by traditional models of research, I mean you recruit participants into the lab, you do in-lab assessment, you put them in an MRI? machine to collect data versus an open science model. In our case it is really just putting the research app on App Store for everyone to use. Now, a second principle that really differentiates between these two different models, in my opinion, is this concept of “return of value.” And by that, I mean not only do we take data from our citizen scientists, we also strive to provide feedback information on the dashboard of our app so that they can potentially learn more about their own brains. For example, they can see how many key presses they have entered for today just like you can see how many steps you’ve taken if you wear a Fitbit.

I always use every opportunity I can find to really express my gratitude towards our citizen scientists. So as of today, we have recruited more than 2000 citizen scientists who really trusted us with their typing data.  and collectively we have obtained more than forty million key presses from them. Again, they really trusted the research team. I always want to make sure I acknowledge their contribution.  And to the best of my knowledge, our BiAffect project is really the first large-scale research effort to look into whether smartphone typing would provide useful information about a person's brain health and mental health. And partly thanks to our BiAffect research effort, I think the past few years there has been growing interest in looking into smartphone sensors. Traditionally people talk more about GPS, location, activity, heart rate, but nowadays, people are starting to realize that keyboard can be another great source of information. And I think part of that is, again, a contribution from our research community. And I want to just really think our citizens scientists for being part of this effort. 

18:20   VO:

Dr. Leow shared some of the advice imparted by trusted mentos and role models that has guided her research career and help, as she puts it, turn science fiction into science.

18:35   LEOW:

Every time I have a chance to introduce our BiAffect community to other researchers, either in conferences or various speaking engagements, I like to frame it as turning science fiction into science.

What I have learned in this process of turning science fiction into science is that it really takes a village of open-minded people. 

There are quite a few folks that I have come across in recent years and who have helped me quite a bit in sort of framing my own career trajectory. And again, coming back to this notion of turning science fiction into science, what are the necessary components or ingredients for something like this to happen? 

A couple years ago, before COVID, I was introduced to Dr. Robert Langer, and he is a distinguished professor at MIT. And nowadays he's probably best known because of COVID and for his co-founding role of Moderna. He is one of the most cited researchers ever in history and he's such an amazing role model for a lot of researchers in drug delivery. After COVID when I was able to connect the dots- no pun intended- I realized that Dr. Langer is probably someone I should talk to for personal advice if I am ever given the chance. So, although it was really just a casual introduction at that time, I finally decided that well, what do I have to lose? Worst case scenario he doesn't reply to my emails. So I took the initiative, I emailed him and he actually started communicating with me and provided me some personal guidance and advice over the past six months. And one thing that he mentioned- and he talked about this all the time, in his talks- he always mentions that it is important to be a good human being. When we are in the middle of something, when we're in the middle of an argument or a deadline, sometimes we lose that perspective. So he constantly talks about this importance of being a good human being. And that really resonated with me.

Another concept that he discussed is this idea for the younger generation- make sure they learn the fundamentals and make sure they learn the basics because the foundation is how we build a solid framework for a person to be able to achieve great things. These are the two things that I constantly think about nowadays if I were asked to comment on to ensure that we have a constant flow of a new generation of research scientists who are going to carry the great American tradition of research of innovation.

Before Steve Jobs passed away, he gave his speech at Stanford- I'm paraphrasing it from memory- but he basically said, “I think when we are in the middle of something, we sometimes are so caught up in this mode of trying to make sense of all the data points we're seeing right now and sometimes that can be anxiety provoking.” 

22:40   VO:

A quick note, I tracked down this speech. If you’re interested, you can find the link in the show notes. Back to Dr. Leow…

22:45   LEOW:

I know for many years I was anxious about making sense of my own career. He said you can't really connect the dots looking forward, you can only connect the dots looking backwards. And so you just have to have faith that, in the future when you look back, all the dots are there. And so what I’ve learned in this process is that as long as I keep an open mind, always be willing to acknowledge what I don't know, and always engage in a good productive conversation with a person even if- or especially when- you don't agree from a scientific perspective, because that's really when new ideas can emerge. And so that's how we could potentially put together a multidisciplinary team and really do amazing research. So, looking back, all the dots were there. I was finally able to reconstruct why we were able to be so lucky to put together the BiAffect research team. But few years ago, when I was starting this process, I was so afraid that this wouldn’t work. 

And what I really appreciate for people nowadays that I get to work with, and I constantly tell myself that I have to be exhibiting the same qualities as I observed in my wonderful collaborators and colleagues, which is this this concept of metacognition, this ability to know what you don't know, to be able to sense that, you know, no matter how firm a belief is in a specific area, there's a possibility that I'm actually very wrong. And so, I should always be having that open mind and be willing to talk to people in an entirely different area and see what they have to say about the specific issue. And I've noticed by doing this, you actually get into a very productive brainstorming session, so to speak, and then great ideas and great discoveries can really come from that.

25:20  Voice Over:

Collaborative Endeavors is produced by me, Lauren Rieger, on behalf of the Center for Clinical and Translational Science (AKA the CCTS) at the University of Illinois Chicago. 

To learn more about BiAffect, Dr. Leow and the research discussed in this episode, visit the links in our show notes.

The CCTS is supported by the National Institutes of Health’s National Center for Advancing Translational Science through their Clinical and Translational Science Award. Opinions expressed by guests of the show are their own and do not necessarily represent the views of myself, the CCTS or our funding agencies.

You can find more episodes of Collaborative Endeavors on Apple Podcasts, Spotify, Amazon Music & Google Podcasts. New episodes drop monthly. 

To learn more about how you can get involved in health research, visit or follow us on Twitter @UIC_CCTS.