{shortcode-9a1a4dc02930fe4cd55603bf827f347f6f9ec86c}

Harvard affiliates developed a silicon chip that successfully mapped more than 70,000 synaptic connections from 2,000 rat neurons, advancing a new recording technology to address existing limitations in neural imaging.

The research from Harvard’s School of Engineering and Applied Sciences was published last week in Nature Biomedical Engineering, a peer-reviewed scientific journal.

“Nobody has been able to do this kind of thing in the recording business,” said Donhee Ham, a Harvard Applied Physics professor. Ham is the principal investigator who led the research group behind the publication.

The most recent publication builds upon a previous neuroimaging project also created by the Ham group in 2020, where they mapped more than 300 synaptic connections. The new silicon chip, which captures 70,000 plausible synaptic connections, enables scientists to read neural activity at a significantly higher resolution and construct a more accurate biological map of brain connectivity.

Researchers previously relied on electron microscopy to generate an anatomical map of neurons. This technique, however, did not allow for determining the strength of connections and deciphering dynamic information.

Though there are other Nobel prize-winning neural recording techniques, they do not measure neuron activity in a parallel fashion. Patch-clamps only record a single neuron’s activity and are not scalable. The current method developed by the Ham group scales up and records signals at higher sensitivity levels.

“This project is about getting mapping from neurons,” Jun Wang, a postdoctoral researcher in the Ham group and first author of the paper alongside Woo-Bin Jung, said. “The mapping means we want to know how they connect — how the neurons are connected — and we want to know the connection strengths of the neurons.”

“We can start seeing how, as neurons grow on these platforms, how they are able to set up and set themselves up so they can survive,” Nadir Talha, a graduate student in the Ham group, said.

Talha said that this would eventually help scientists understand how “animals, or maybe even humans, can configure their networks to be able to perform higher tasks.

“How the neurons connect and generate function is the central importance in the neuroscience field, or even the artificial intelligence field,” Wang said.

At a moment of heightened interest in silicon chips and microprocessors that function like brains, the research marks a breakthrough development in neuromorphic engineering and AI.

Wang said he hoped the research would help scientists develop the “next generation of artificial intelligence systems.”

“We delivered a tool that may help give other neuroscientists, or other researchers, a new platform to study their questions,” he added.

Ham said he was “very proud of Jun and Woo-Bing.”

“At the same time, it’s a really small step, because the number of synapses that the brain has is humongous,” he added.

—Staff writer Xinni (Sunshine) Chen can be reached at sunshine.chen@thecrimson.com. Follow her on X @sunshine_cxn.

—Staff Writer Danielle J. Im can be reached at danielle.im@thecrimson.com.