The scanning and computational reconstruction were conducted by Canadian and German researchers, using the brain of a 65 year old woman who donated her body to science. The project was part of Europe's Human Brain Project. Over thousands of hours, the brain was thinly sliced, scanned, and then reconstructed as a virtual 3D model that is freely available for scientists to access.
The next important step in this work, as I see it, is for the 3D physical model to be turned into some form of computational model like a neural network simulation, either wholly or in part (i.e., a connectome). Then simulated inputs can be fed into the system to see how activity occurs, and to conduct experimental work on this computational model derived from the detailed physical model. To build the connectome, realistic assumptions will have to be made about how the physical connectivity translates into computational connectivity, but this does not seem too insurmountable now that we have a full cellular connectivity map in 3D.
Another issue with the computational reconstruction will be running the simulations on a powerful enough computer system. The computational complexity will be absolutely immense, even with state-of-the-art technology. So perhaps we will be limited for the time being in only simulating certain brain circuits or maybe even regions, but this should be very enlightening in any case.
For further reading on this topic, check out the field of Connectomics, which is spear-heading this type of research. The 1990's was the Decade of the Brain. It looks like these new results may lay the groundwork for a Century of the Mind, in which our understanding of how brains give rise to minds advances exponentially. Perhaps, in our lifetimes, we may see satisfying scientific explanations for mental illness, brain dysfunction, intelligence, and maybe even the mysterious problem of consciousness. Only time (and lots of future research) will tell.
The next important step in this work, as I see it, is for the 3D physical model to be turned into some form of computational model like a neural network simulation, either wholly or in part (i.e., a connectome). Then simulated inputs can be fed into the system to see how activity occurs, and to conduct experimental work on this computational model derived from the detailed physical model. To build the connectome, realistic assumptions will have to be made about how the physical connectivity translates into computational connectivity, but this does not seem too insurmountable now that we have a full cellular connectivity map in 3D.
Another issue with the computational reconstruction will be running the simulations on a powerful enough computer system. The computational complexity will be absolutely immense, even with state-of-the-art technology. So perhaps we will be limited for the time being in only simulating certain brain circuits or maybe even regions, but this should be very enlightening in any case.
For further reading on this topic, check out the field of Connectomics, which is spear-heading this type of research. The 1990's was the Decade of the Brain. It looks like these new results may lay the groundwork for a Century of the Mind, in which our understanding of how brains give rise to minds advances exponentially. Perhaps, in our lifetimes, we may see satisfying scientific explanations for mental illness, brain dysfunction, intelligence, and maybe even the mysterious problem of consciousness. Only time (and lots of future research) will tell.
No comments:
Post a Comment