Interview with Dr. Sarah Genon on a new approach to discover “operational functions” of brain areas
Seeing doesn’t necessarily mean understanding. This brief notion is perhaps the best way of describing the problem that drives many researchers in the field of neuroscience. When imaging techniques such as functional magnetic resonance imaging emerged in the 1990s, it appeared to be just a matter of time until we understood how speech is processed, sentences formed, and recollections stored in our short- and long-term memories. However, the current estimations of many scientists paint a much more sober picture. To date, hardly any concept from the fields of psychology, philosophy, or sociology can be clearly assigned to biological processes and structures in the brain.
Neuroscientist Dr. Sarah Genon, who conducts research at Forschungszentrum Jülich and University Hospital Düsseldorf, even speaks of a “conceptual chaos”. Within the European Human Brain Project, she heads a subproject concerned with the multimodal comparison of brain maps. Together with Prof. Simon Eickhoff, Prof. Katrin Amunts and other neuroscientists from Forschungszentrum Jülich and University Hospital Düsseldorf, Genon is proposing a new approach that could enable the analysis of large data sets and help to considerably further this area of research in the long run.
You had been active in the field of Alzheimer’s research until recently switching your focus to more fundamental questions, focusing on functional neuroanatomy and methods development. What sparked your interest in this field?
During my PhD, I investigated the neural correlates of cognitive dysfunctions that can be observed in patients with Alzheimer’s disease. I used many different brain-behavior approaches and statistical analyses. When I started analyzing the data, it was clear from their cognitive performance that the patients had severe deficits in many memory-related processes. When relating these cognitive data to the brain data across the different studies, I saw very complex patterns. Put it simply, there was not a clear, straightforward relationship between one type of behavioral dysfunction evidenced by cognitive paradigms and one pattern of brain alterations. I therefore decided to take one conceptual step back by trying to understand how the brain works, what it does and how we should study the relationship between brain and behavior.
Technological progress is enabling an increasingly closer look into the human brain. At the same time, there has been continued criticism that research hasn’t made any significant strides on many issues. So what are the problems?
One big question remains open: what does the brain do and how does it do it? Several researchers’ communities study the structure of the brain. After more than one century of work to characterize brain anatomy, there is strong evidence that different regions have different microarchitectures and are organized into networks. Accordingly, the common view is that brain regions have relatively specialized functions but also work together exchanging information. The follow-up question is therefore: how does it relate to human behavior? This is something being studied by many different fields, mainly psychological sciences. By observing and testing human behavior, several theories and concepts have been developed such as “phonological lexicon”, “remembering” or “theory of mind”. Accordingly, across the past decades, neuroscientists and clinicians have tried to relate these concepts to brain regions and networks, but no clear picture has emerged. In other words, behavioral concepts do not map to brain regions and networks. One reason is that there is a conceptual gap between the brain data and the behavioral data. For example, even if we can relate the hippocampus (a part of the brain) to behavioral concepts such as “remembering” or “future imagination and planning”, this is not what the hippocampus does. In other words, those concepts, especially taken in isolation, do not really inform us on the function of a part of the brain.
You are proposing a new approach intended to create order amongst the chaos. How exactly will this work?
First, we have to collect a maximum of information. So far, we mainly had individual neuroimaging experiments in which, for example, an activation of the hippocampus was reported during the recollection of a memory. We now have big collections of such experiments, in such a way that we can run statistical approaches identifying the range of behavioral functions in which the hippocampus has been found to be engaged. We can see, for example, that in addition to being engaged in memory, the hippocampus can be related to perception and emotions. We also now have big datasets including neuroimaging data and psychometric data of large population samples. For a given region, we can therefore search for correlations between neuroimaging markers (such as grey matter) and complex behavioral constructs such as “creative thinking”. In other words, we can conduct different types of “behavioral profiling” of brain regions and networks from a large collection of brain-behavior data to disclose the whole picture. However, this behavioral profiling is just a first step to starting a new conceptualization of brain functions. We also have to extract “latent patterns” and meaning from these data. This can only be achieved by statistical learning approaches and cross-disciplinary theoretical work.
You are aiming to decode certain “core jobs” of brain regions. How are these core functions different from traditional cognitive functions such as speech, memory, and attention control?
Any part of the brain receives inputs and sends outputs from and to other parts of the brain. We could think of the brain as a computer, a statistical machine or an information processing system. For example, we have to think that the brain combines information to compute probabilities and make predictions. Function “in the brain view” should therefore be considered as some type of “computation” based on some “representations” or “information”. In contrast, “language”, “memory” or “attention” are by essence human (or animal) functions. They refer to the human behavior that we can observe in everyday life. The type of functions that the brain performs, in contrast, cannot be directly observed.
What exactly do you imagine a core function of this kind to be like? Can you maybe provide an example?
Nobody knows exactly what it should look like. The first question is: “Is the basic function of brain regions only some type of integration of different information or are there different types of operation that are performed by different brain regions?” Given that different parts of the brain, such as the frontal cortex and the hippocampus, have different microarchitectures, we could expect that they have different computational properties. An approximation of the type of function that the hippocampus performs could be, for example, a type of “high-resolution binding” (cfr Yonelinas, 2013), in other words to link specific high-resolution features, in service of cognition functions from perception to memory.
How can research benefit from an orientation towards core functions in the long term?
Understanding the job of any brain region is fundamental to understanding neurocognitive dysfunction in patients in neurology and psychiatry. We can observe relatively specific cognitive dysfunction in patients with specific brain lesions and certain brain pathologies. However, the mechanisms from brain dysfunction to behavioral dysfunction remain relatively poorly understood. There are also presumably some specific dysfunctions that are not clearly evidenced because they do not result in obvious perturbations in the behavior of the patients, but they could be predicted from a better understanding of the function of brain regions and networks. Thus, understanding brain function goes hand in hand with understanding neurocognitive dysfunction.
Despite all the problems imaging techniques as functional magnetic resonance imaging offer great new opportunities. To what extent can imaging techniques help to provide a better understanding of how humans feel, think, and act? What important successes have already been achieved?
Some important findings in our research recently have been to provide maps of brain organization, highlighting differences between brain hemispheres and inter-individual variability. The general pattern of brain organization, its variability across individuals and how to relate to difference in behavior and cognition is a major topic for future research. We have for example already shown that a similar behavioral measure can be related differently to brain features in men and women, or that brain aging does not strictly follow biological aging. This means, for example, that some people might be 70 years old but their brains look like that of a 60 year-old in term of atrophy. These findings are crucial as they demonstrate the flexibility and the plasticity of the human brain and they are fundamental to developing adequate treatments of dysfunction for everyone in the future.
Is there an area of the brain or a core function that you are personally especially interested in?
I am particularly interested in the hippocampus and the frontal lobe. I initially did not want to study the hippocampus simply because I was aware that thousands of researchers have been trying to decode this little structure that keeps on hiding its secret and, therefore, challenging many research communities. However, every scientific or philosophical question I have thought of, always puts the hippocampus back in the picture, from “how does the brain encode information” to “how do we build and update self-related beliefs?” Finally, I have seen the beauty of the hippocampus in the eyes of a neuro-anatomist looking at a brain section. From his eyes, it seems like he was looking at one of the wonders of the world.
Dr. Sarah Genon, many thanks for this interesting interview!
More information is available on our website.