Q: You concentrated on theoretical solid-state physics in post-graduate school, right?
A: Yes, My work had nothing to do with neuroscience; however, theory can generally be applied across fields. After graduating, I plowed through various ideas in physics in search of my next research subject; the field is certainly generous. It provides its researchers with many interesting areas to explore, but seemed removed from actual social connections. I wanted to attack a subject that might also interest, or convince, non-scientists too. Therefore, I became a researcher in the visual model research group at NEC Fundamental Research Laboratories (now NEC Fundamental and Environmental Research Laboratories). My longstanding, if vague, interest in how my brain works may also have played a part in this decision. Back then I had no knowledge about the field, I could not even explain what a neuron was, so I forced to struggle each day. After a full day at the laboratory that started at 8:30, I would return home to study my newly chosen subject until 3 or 4 o’clock the following morning.
Q: How was it that you selected your current research subject?
A: When I noticed that in the primate primary visual cortex, columnar structures of neurons sharing similar response properties are arranged regularly, I realized that these structures resembled the domain structure in a ferromagnetic thin film. This thin film has a stripe pattern composed of domains magnetized upward or downward, perpendicular to the film. When an upward magnetic field is applied to the film, the upward magnetized bands expand and instead the downward magnetized bands are narrowed. Such changes in the magnetic domains in response to the external magnetic field are quite analogous to the changes in the widths of eye-dominance columns induced by the deprivation of one eye during development. Then I started to build a viable mathematical model of column formation in the visual cortex, using a theoretical framework of the domain structure formation in the ferromagnetic thin film. I was amazed that the brain, which had previously appeared desperately complex, had an orderly structure, and sure that a viewpoint of physics I had learned would also be useful in the study of the brain. I had felt as if I was standing in front of a precipitous cliff so far, but finally I could find a foothold for climbing.
Q: Your research on the brain was theoretically based?
A: Initially yes. But, as I began writing papers on the phenomenon of self-organization, my father contracted hydrocephalus due to poor cerebrospinal fluid flow following an accident. At that time I realized how powerless I was as a researcher of brain theory. I felt that I should research the brain in ways that draw from or contribute to real experience. It was at this time that RIKEN invited me to join them and conduct actual experiments. These experiments investigated the visual cortical columns in cats using optical recording. I wanted to verify self-organization phenomena suggested by my model experimentally, thereby providing a deeper understanding of these regions. Nevertheless, I lacked experimental experience. I has also never seen a brain. I hired a post-doctoral researcher with experimental experience to begin experiments, but this researcher left RIKEN a year later. And, the student who took over, broke his arm. In the end, I had to carry out the experiments myself. Having worked exclusively in theoretical areas, experimental research was yet another giant leap in the dark for me. My first experiments were completed relatively easily, thanks largely to tutorial instructions from my student, who was standing behind me. I have continued to conduct experiments by myself, I even practice image training for experiments wherever I can. Including on commuter trains. This approach was very helpful in executing my experiments smoothly. Image training has proven itself to me.
Q: What kinds of experiments have you carried out?
A: For one experiment, a cat was given only vertical stripes as visual stimuli for a period of time. We then checked if the columns responding to the vertical stripes. Did they really expand as expected from self-organization theory? Many papers denied such a phenomenon, but I was convinced from a theoretical viewpoint of my conclusions and doubted that the experimental methods were in fact appropriate in those papers. So I visited a do-it-yourself shop. I wanted to find a way to fabricate cat-size goggles. I had test cats wear the goggles permanently and gave them only visual experiences of the outside-world whose images were elongated in a single orientation. The results of optical recording matched what was predicted by our theory.
Q: In brain research, are there approaches that are unique to a physicist?
A: Physics has taught me that all natural phenomena, including those of the brain, should be explained using as simple a model as possible. In fact, it is only after a complicated phenomenon has been successfully replaced with a simple proposition that we can say that we really “understand” the phenomenon. Yes, biological phenomena are diverse and complicated, due to the large number of uncontrollable parameters and individual differences. But well-ordered rules can always be extracted from such diversity. My way of thinking is imprinted with the theoretical research methodology in solid-state physics. I cannot conduct any theoretical research that ignores measured phenomena. It is a hard path I walk, but my research is committed to developing theoretically persuasive understanding of the brain.
