Harvard University affiliates have developed a new technology for mapping neural connections to understand how the brain works.
Harvard University affiliates have developed a new technology for mapping neural connections to understand how the brain works.
A silicon chip developed by Harvard University researchers has successfully mapped more than 70,000 synaptic connections from 2,000 rat neurons and improved a new recording technology to overcome existing limitations in neural imaging, The Harvard Crimson reported, citing research from the Harvard School of Engineering and Applied Sciences published last week in the peer-reviewed journal Nature Biomedical Engineering. The results allow scientists to read neural activity with much higher resolution and create a more accurate biological map of the brain’s connections.
The principal investigator who led the research team, Harvard University professor of applied physics Donhi Ham, said that the development is based on a previous neuroimaging project, also created by Ham’s group in 2020, where they mapped more than 300 synaptic connections.
The advantage of the development of Harvard scientists over other methods, even those awarded the Nobel Prize, is the ability to make maps of neurons, learn «the strength of neuronal connections» and «begin to observe how neurons grow on these platforms, how they are able to adjust and tune themselves so that they can survive,» the group’s researchers report.
Nadir Talha, a graduate student in Hema’s group, said that this will ultimately help scientists understand how «animals, or perhaps even humans, can tune their networks to be able to perform higher tasks.»
The research marks a breakthrough in neuromorphic engineering and AI because it sheds light on how neurons connect and generate functions. And according to Jun Wang, a postdoctoral researcher in Ham’s group and first author of the paper along with Woo-Bin Jung, this is central to the field of neuroscience or even the field of artificial intelligence.
He hopes the research will help scientists develop «next-generation artificial intelligence systems.»
