“To me, it seemed like building something — whether it’s an app, or an actual tool, or a device that someone uses — was a more concrete goal that I could literally put my hands on.”
Elena Allen, Neuroscience PhD Program alum (entering class of 2003)
Elena Allen, chief scientist at Rodin Scientific, is developing products to allow patients to monitor their chronic heart failure at home, reducing the need for repeated trips to the hospital. Her company is developing two noninvasive products — a wearable ring and a box that goes around the hand — that patients can use to assess their condition at home. If there is a problem, they can adjust their medication or diet accordingly, before a serious health crisis occurs.
Allen is a Berkeley Neuroscience PhD Program alum whose expertise is rooted in data science and analytics. As an undergraduate at UCLA, she was an applied mathematics major. She pursued a PhD in neuroscience because she was fascinated with biology and the brain. At Berkeley, she studied visual pathways in the brain with Ralph Freeman (professor emeritus, optometry and vision science), with a focus on the technologies of functional MRI and transcranial magnetic stimulation. After graduating, she did a postdoc at the Mind Research Network in Albuquerque, New Mexico and a second postdoc at the University of Bergen, Norway. Allen then went into industry, first through consulting, and then as a senior scientist at Medici Technologies which later launched Rodin Scientific.
Read our Q&A with Allen to learn about her career path, experiences at Berkeley, and the excitement she feels when she sees her ideas turn into tangible products. This interview has been edited for length and clarity.
Rachel Henderson: As an undergraduate, you were an applied mathematics major. What led you to do a PhD in neuroscience?
Elena Allen: That became a plan around my junior year of college. I wanted to do something in biology, particularly the brain, and at the same time I loved my math classes. In talking with some professors from the neuroscience department, they made the argument that being very comfortable with the quantitative side of things would give me maybe an edge, or at least a lot of comfort in studying neuroscience, because increasingly, it was becoming a highly quantitative domain. The idea was that being comfortable with more analytical tools would at least give me more options in future studies. So I took a lot of neuroscience courses, the full complement, I think, of what was offered to undergraduates at UCLA at the time. I really enjoyed them and knew that I wanted to do more there. But I stuck with the math degree just because it was so much fun!
RH: Why did you end up coming to the Berkeley Neuroscience PhD Program?
EA: Many reasons. My family lives in Southern California, and there’s something nice about being distant, but close. During the interview process, I really liked the graduate students who were in the classes above me. You also see some of the same people you are interviewing with again and again, and I knew a lot of them were thinking about Berkeley. So a confluence of factors, I think. People seemed genuinely happy with the HWNI [Helen Wills Neuroscience Institute] program, and I think I kind of vibed with the general personalities of the other graduate students there. There was potentially more of a cultural dissonance at the East Coast schools, just because that’s like a really foreign environment to me. They felt more formal, and Berkeley just felt comfortable.
RH: What was your PhD thesis about?
EA: It was about neural and metabolic processing in the central visual pathway. I worked in Ralph Freeman’s lab in the School of Optometry. In the lab, there were experiments on the thalamus, the LGN [lateral geniculate nucleus of the brain], and the primary visual cortex.
The lab did all sorts of things. There was some really interesting work that had been done in the lab by Jeff Thompson, prior to when I got there. He was looking at the signal contributing to the BOLD response [Ed. note: a measure of blood oxygenation detected by functional MRI that is an indicator of brain activity] in the micro sense, using a custom oxygen sensor that looked at oxygenation changes in the surrounding tissue. The elusive initial dip in oxygenation (that maybe you could see in neuroimaging or maybe not, depending on a variety of different parameters) was very evident at that small scale, before the large positive level of response that came subsequently. So I carried on with some of that work.
I also did another project that was largely due to the influence of Brian Pasley, who was a fellow graduate student in my Helen Wills [Neuroscience Institute] class and also in Ralph’s lab. He was doing a lot of work in humans and was interested in TMS (transcranial magnetic stimulation). There was kind of a dearth of knowledge in terms of what that actually did to neurons. So we started doing some combined experiments using all sorts of tools like optical imaging, single-cell recordings, local field potentials and looking at neural responses to magnetic stimulation in the cortex, and then the potential effect on hemodynamic coupling.
I would not say that I had a very clear vision of what I wanted to research [laughs]. I was put in a really high quality research environment, and I looked at what other people were doing and either found it interesting or didn’t. It was like a buffet, and I chose things that I thought would be delicious. Whether or not they actually turned out to be delicious, I would find out, but I didn’t have a plan going in, that’s for sure.
RH: What was your overall experience like as a grad student at Berkeley?
EA: It was really fun. I lived in a house owned by another Neuroscience graduate student, Annaliese Beery and her partner (now husband), and it was just full of graduate students. So my social world involved a lot of the other Helen Wills folks. I found it a really enjoyable atmosphere. The program really worked to support us with a lot of good activities. Overall, it was a very positive experience.
Looking back, I think I could have had a lot more fun had I been less afraid of my fellow students. I had huge imposter syndrome that was nearly crippling. So I probably excluded myself from a lot of things that I would have really benefited from or enjoyed, just because I was so impressed and I felt very out of my league. There’s just so many smart people, that you can feel alone. It can be daunting, sometimes, even just to hang out with people and drink a beer [laughs]. I probably could have had an even better time had I been able to get past that.
RH: You did a couple of postdoctoral fellowships after you graduated — what did you do for those?
EA: When I first graduated, I didn’t know what I wanted to do. I felt pretty burned out on research. I spent a tiny bit of time, actually while I was writing my thesis, teaching in a high school which was really, really satisfying. I enjoyed that a lot. Then my partner, now wife, got a job. She had finished her PhD a year earlier and got a job at a research lab in Albuquerque, New Mexico. Neither of us were determined to stay in the Bay Area, we wanted to check out other things, so we moved out to New Mexico. I started looking around at opportunities there, and I ended up accidentally pursuing a postdoc [laughs].
I started working with Vince Calhoun at the Mind Research Network (a private organization associated with the public university here in Albuquerque), focusing on neuroimaging. His bent was very much big data, analyzing neuroimaging data from hundreds of people, and using data-driven methods to extract information. That was such a productive environment. I was in that lab for about three years. There were so many good people from all around the world. I made collaborators and friends there who I still consult with. I’ll call up and say, ‘I don’t understand this statistical plot, can you explain it?’ That is just New Mexico’s vibe; it’s casual. Even way, way more informal and relaxed than Berkeley, and that was a really nice fit for me.
In that experience, I met collaborator Tom Eichele who then started his own lab at the University of Bergen in Norway, and I did a postdoc with him. That was a good experience, much more focused on the application of methods developed at the Mind Research Network. There was a big adjustment though, from being in an environment where I had an entire network of collaborators one desk away, to one where I was more independent. Then my dad got sick with brain cancer, so I moved back to the States. That was a difficult couple of years, and I ended up finishing up the postdoc remotely so I could help with family stuff.
RH: What led you to go into industry after that?
EA: The things that keep me motivated on a daily basis are the people that I work with, and the end goal. Everyone has their own driving forces, but for me, if I’ve got both of those, I can almost always feel motivated to get things done. I think what’s challenging in research is that sometimes the end goal is really, really far away — decades, lifetimes away. I think that was a real struggle for me in research. The end goal often seemed to be the prestigious publication. That was a convenient placeholder, because the ultimate goal of understanding how the brain works was so daunting and so far away. After five years of grad school and then a couple postdocs, I was like, I don’t think I care anymore about publishing. I feel like that’s not enough for me. Obviously, scientists aren’t in it for the publishing, but I think it often becomes a substitute short-term goal — you have to build out your CV so you can get funding, so that you can have a job.
So I knew I wanted to try out something different from pure research and I started looking at doing some consulting while finishing up my postdoc. At this point, the whole idea of a data scientist was really growing. People were looking for individuals with skill sets that were quite harmonious with what you do when you’re studying neuroscience. You’re doing research, you’re cleaning data, organizing data, building things, testing models, and doing machine learning. That skill set had a high degree of overlap with data science. So I started doing some serious projects, and ended up doing consulting for an individual who was trying to start his own data science company. At some point, I transitioned from being a consultant to an employee, and then that company spun off into another company.
The goals of both these companies were dual in nature. One, to make money, because it’s a business and it needs to support itself. But more importantly, because the individual who established them is a former MD and an engineer, they were designed with the purpose of solving problems in healthcare. Whether from a data analytics perspective, or from a device perspective, or from a patient engagement perspective; these were all kinds of things that we were discussing. To me, it seemed like building something — whether it’s an app, or an actual tool, or a device that someone uses — was a more concrete goal that I could literally put my hands on.
RH: How was the transition from academic research into industry for you?
EA: I like to learn anything new and I think I am truly not unique in that. I think most people pursuing any sort of higher degree are often driven by learning new things. The transition to industry provided a lot of new things to learn, which was really motivating and exciting. For example, writing a patent is sort of like a negative inversion of the process you use to write a scientific publication. When writing a publication, you’re like, okay, we need to be clear about what we did. Here are the steps that we implemented so that others can replicate it. Writing a patent, you want to write about any way this could be done — ever. You have to write it before you even build it in many cases, as soon as you think it’s worthwhile. You have to project into the future or into different domains and think, what are all the ways you could do this? What are all the ways we could build it? And can I write a patent that protects all of that? It’s this crazy intellectual exercise that I find totally exhausting. But it’s been fun to learn to do that; it’s something totally different from my prior training.
Also, building things, working with engineers — that’s another really satisfying aspect of being in industry. And talking to design firms about: How would we put this together? What are the human factors that we need to think about? Those are not things that I learned in graduate school or in prior research. In more academic environments, you’re kind of just like, ‘Can we build it with duct tape and $50?’ and that’s the goal. In industry, when we’re first starting out, it’s all duct tape. But then to see it move out of that phase is so exciting. One of our devices had this old car carburetor on it, and to see it move from that to something that a design firm put together, and then have it built in China — that’s an amazing feeling. So I think learning new things is exciting. You do it in grad school and you do it as you move to this domain as well.
RH: What does your company (Rodin Scientific) do, and what is your role?
EA: I do research here. We’re trying to build home-based devices to help individuals manage heart failure. Heart failure is really broadly defined as anytime your heart can’t pump enough blood around to meet the needs of your body. You can develop it for a whole bunch of reasons — it can be that you’ve endured 30 years of hypertension and your vasculature isn’t compliant anymore; or you’ve had a heart attack and your heart muscles weaken; or you’re born with a defect in your heart, so it just doesn’t pump well enough. Right now, heart failure is the most expensive chronic condition to treat in the US. It costs upwards of $40 billion a year, and that’s almost all public money because the vast majority of heart failure patients are covered by Medicare. So this is a significant drain on our healthcare system.
But the cost of heart failure treatment is a problem you can do something about, because about 80% of the costs are related to hospitalization. What tends to happen is that patients have some sort of exacerbation — they might get a cold that causes some buildup of fluid in their lungs, or they might go off their low-salt diets and retain extra water. And because the health of these patients is always on the edge, really precarious, these little exacerbations can cause serious decompensation. One of the most dangerous consequences is severe fluid buildup in the lungs, so that patients can’t breathe, sending them to the emergency room. Patients can have multiple hospitalizations per year in some cases, which is a huge burden on them and on the healthcare system.
So the question is, can we build a home-based device, similar to in diabetes where we have individuals tracking their blood sugar levels on a daily basis, and then proactively manage patients to help them avoid an ER visit and hospitalization. If we have the right tools in the home, can we track these individuals and proactively make changes in diet and medication to keep them out of the hospital? That’s the problem we’re trying to solve with the company.
There are many other companies out there trying to solve this problem, but we’re determined to do it with completely noninvasive technology. There are several existing devices that are invasive — you have to stick them inside people. But obviously, not everyone wants to have devices stuck inside of them. If you can get the same information noninvasively, it’s cheaper and much easier on the patient. And you can start to treat a patient population that is not as severe as the group that is well-suited for a highly invasive device, potentially slowing or preventing disease progression.
RH: Do you have products that people are using now?
EA: We’re not quite at the product stage, we’re at the research and development stage. We’ve done about eight different studies, testing out the tools, validating the measurements, and improving the design of the instruments to get the best quality signal that we can get.
We have two different tools. One is a wearable ring. The base technology is similar to what’s used in an Apple Watch — you’re using light to measure blood volume, and one of the most prominent changes in blood volume is the pulse. So we apply algorithms and models to the pulsatile blood volume signal to extract information about the stroke volume — the output the heart is producing.
Another device we make is a little box you put your hand in. We use optical imaging to take pictures of the veins in your hand and change the pressure around your hand to see how it affects the veins. From this, we determine the pressure in your veins, and then extrapolate your central venous pressure. Your central venous pressure is the pressure going into the right side of your heart. Oftentimes, in patients with heart failure who are becoming fluid overloaded, the central venous pressure will become elevated. The pressure of the blood returning to their heart is too high, which is definitely something you would want to treat. Currently, there’s no tool available to patients that can measure and monitor their central venous pressure, with the exception of inserting a catheter into a jugular vein, which doesn’t work so well in the home.
RH: How do you think your experiences as a grad student at Berkeley have helped you in your current career?
EA: I think it’s just the caliber of individuals there, in a lot of different domains. It could be anything from their analytic prowess to introducing you to new ideas. It’s eye-opening to watch someone who is an incredible communicator, seeing them deliver a presentation and being like, ‘Whoa, you’re really good at that. I want to get good at that too.’ I think that talent, in all domains that you can find at Berkeley, inspires you to hold yourself to a higher standard, and encourages you to learn new things.