In the lead up to the OHBM Annual Meeting, I had the pleasure of speaking to one of the keynote speakers, Dr. Biyu He, an Assistant Professor at New York University. Dr. He has made many valuable contributions to the field of neuroscience, combining diverse imaging methods and analytical techniques to tackle big questions relating to perceptual processing, spontaneous activity and consciousness in the human brain.
Rachael Stickland (RS): Thanks again for joining me. It's nice to - virtually - meet you.
Biyu He (BH): Pleasure to meet you as well.
RS: I'm getting used to having many video calls every day now. I'm sure you are as well. How have recent months been for you, adapting to working remotely and only connecting to most people virtually?
BH: It's been okay. I miss the face to face interactions with people. But I think we've been very adaptive in my lab. As you know, in human brain imaging, we do a lot of data analysis. So we have been working on reading, writing and data analysis. And I think we've been able to weather the strange situation we live in pretty well.
RS: You're currently based at New York University (NYU) as an Assistant Professor in the Departments of Neurology, Neuroscience & Physiology and Radiology. Do you mind telling me about your research path and your route into science?
BH: Sure. I was a biology major in college, and really liked maths and physics when I was young. I wasn't sure what I was going to do in college initially but once I found neuroscience I was immediately hooked. It is just so absolutely fascinating. I felt like I couldn't ever be bored again. And it's also one of the most interdisciplinary fields in science. It's challenging and fascinating and very, very intellectually engaging. I did my PhD in neuroscience at Washington University in St. Louis. From there, I was looking for postdoc positions at the end of my PhD and unexpectedly got offered two positions to set up my own lab. One at the National Institutes of Health (NIH) and one at the University of Konstanz in Germany. I decided to go to the NIH and spent about five and a half years there. It was a wonderful time — I learnt new techniques, made new friends, found new mentors, and started a new line of research, which is what I'll be talking about in my [OHBM] keynote talk. Then, I moved to NYU a few years ago.
RS: You mentioned how neuroscience is very interdisciplinary. That might be why it’s hard to explain what we do! If a non-scientist asked you what your research is about, and also why it's important, what would you say?
BH: Broadly speaking, I’m trying to understand how the human brain generates conscious awareness and conscious experiences. And how neural mechanisms underlying conscious awareness differ from, and interact with, unconscious processing. From decades of research in psychology, we know that sensory input impinging on the brain can be processed by the brain consciously, giving rise to all the experiences that we enjoy, but also unconsciously. So things that you don't consciously perceive can nevertheless influence your behavior. We don't really know what neural mechanism gives rise to conscious experience and how that differs from unconscious processing. Understanding the neural underpinnings of these processes and their differences is very important for a lot of clinically and societally important questions. For example, we'll be able to better treat disorders of consciousness, including minimally conscious states and vegetative states, as well as many clinical conditions with disordered perceptual awareness, such as hallucinations in schizophrenia, tunnel vision in autism. These are cases where you have disturbed conscious perception. In addition to applications in the clinical and societal domains, addressing this question also satisfies a fundamental human curiosity that is ‘Who are we? Why are we sentient beings? How are we different from robots?’
RS: That’s fascinating. I think scientists and nonscientists alike find the topic of consciousness very interesting. So do you think that fMRI has a key role in helping us understand consciousness?
BH: Absolutely. It's the best method for non-invasively measuring whole brain activity and finding out where in the brain some type of information is. In my mind, it is especially powerful when we combine fMRI with other techniques with higher temporal resolution, like MEG, ECoG or EEG. In human brain imaging, we have a lot of complementary techniques that are very powerful and can give us a view of whole brain activity or large-scale brain network activity, which you could say some of the more traditional animal research techniques haven't been able to get at. But, obviously, there's a lot of push to do large-scale simultaneous recording of many neurons and across many brain areas in animal models now as well.
RS: So your own research combines many of these techniques you just mentioned - invasive and non-invasive methods of studying the brain, including many different human neuroimaging methods. What are the main challenges with integrating such diverse methods, in terms of the experiments themselves but also in the interpretation of findings?
BH: Probably the main challenge is to grasp a lot of literature that's grounded in different techniques, because, when I was a PhD student, I realized that for the same question there is parallel literature, depending on if you use fMRI or EEG/MEG and then the insights are different. The questions and the debates people care about are also different. Each technique is like a window into the brain with its own vantage point. So if you only look through that one window, your field of view is somewhat limited. When combined, the knowledge and the insights from multiple techniques to understand the same biological question can provide a much broader view and you can get at the mechanisms better. Ultimately, we want to understand the mechanisms of how something works in a computational sense: how do neural circuits do the information transformation that allows certain perception and cognition to happen. And for that reason, simply mapping where or when would not be sufficient. We need to combine the insights from these different angles to build a full answer that addresses the mechanisms.
RS: Yeah, that makes sense. So, non-neuroscientists may be surprised just how much our prior knowledge and experience can shape how we perceive something in the present moment, and your research has advanced the scientific understanding on this topic. Related to that, what scientific finding have you found most surprising in your career? Has there been something that particularly surprised you about the brain?
BH: What you just mentioned was a finding that was actually very surprising to me. Me and my lab, when we made the discovery, we actually literally scratched our heads for several months before things started to make sense. You're absolutely right that past experiences and prior knowledge have a profound impact on perception. And it's very interesting because there are certain clinical disorders, including schizophrenia, autism, PTSD, where we know that this process is abnormal. There has been a lot of behavioral and neuroscience research done on this topic. What was really surprising in our findings was the spatial extent of the prior knowledge's impact on perceptually relevant processing across the brain. It used to be thought that visual perception, for example, is basically solved by visual regions. But what we found was that when you go to the really higher-order regions in the brain, even the so-called default network (that is the most remote from sensory input and the apex of the cortical hierarchy) they are involved in this process of prior knowledge guiding visual perception. It's not just that their activation magnitude changes, but their activation pattern changes as well. The voxel-wise activity pattern in those regions reflected the content of prior knowledge and the content of perception. So, that was very surprising. I think, in retrospect, it made sense because this process of prior knowledge guiding perception really requires many different brain networks to work together, from those processing sensory input to those mediating memories. We are still working on the exact mechanisms involved in this. But in the broader picture, it suggests that in real world vision, real world perception, where past experiences continually guide our perception, much more of the brain might be involved than we initially thought.
RS: Your research has brought new insights into the best ways to measure, categorize and model brain activity. Moving forward, what do you think are the most important questions that need addressing, or the most important technological advances, in order to progress understanding in your field?
BH: I have two thoughts here — one one is broader than the other one. The first one is that we need to integrate resting state approaches and task-evoked approaches. There's a huge amount of insight that has been learned, and to be learned, from both approaches. But each approach alone obviously won't be able to resolve how the brain works. I think we have made a lot of progress with both of those approaches, but exactly how we integrate the insights and their analysis methods, that is something that has a lot of room to be developed in the coming years. For example, related to my research topic, conscious perception: I don't think a system without spontaneous activity will have conscious perception; I think it will solve perceptual tasks, but it will not have perceptual awareness. Currently, we have a wonderful, beautiful field of knowledge based on resting-state studies but there is a gap between these insights and what we know about the neural mechanisms underlying perception and cognition. I think at the junction between those two fields, there is a lot of progress to be made.
And the second is something that I alluded to earlier (I think this is where the field is already going), which is to go beyond the mapping of where and when to get at the computational mechanisms. And there are many different ways of getting at the mechanisms — it probably requires leveraging multi-facetted analysis techniques to understand exactly the computational mechanisms as embodied in neural circuits and networks that underlie perception and cognition.
RS: What was the best piece of scientific or career advice you've received? What has helped you to get to the position you are in, carrying out brilliant research?
BH: Thank you. Something that comes to mind is when I was doing my PhD, my PhD advisor, Marcus Raichle, often told us that “Science must be done for its own sake, for any other harvest is uncertain.” It is important to enjoy the science you do. If not, you probably should do something different. That advice has propelled us to pursue questions we are passionate about.
RS: Your OHBM keynote talk is titled “From Resting State to Conscious Perception”. Can you give us a teaser or sneak preview of some of the interesting topics you will cover?
BH: It’s kind of a personal journey of how my scientific career has evolved, and how my work continues to make connections between these two areas. As you can see, from what I alluded to earlier, I think understanding the neural basis of conscious perception requires us to take into account the role of spontaneous brain activity and past experiences that persist through the resting brain. I've been to OHBM almost every year since I was a student, so it's very gratifying for me to be able to tell this personal journey through the different scientific questions I've investigated.
RS: Well, that's great. I look forward to tuning in and hearing it online.