Artificial intelligence (AI) networks designed to mimic the human brain are able to perform tasks more efficiently than other systems, a new study has found.
Researchers from The Neuro (Montreal Neurological Institute-Hospital) and the Quebec Artificial Intelligence Institute developed an artificial neural network (ANN) computing system based on a biological brain’s connectivity.
Training the system to perform a memory task resulted in the ANN completing it more flexibly and efficiently than more traditional systems that do not reflect the organisation of real brain networks, they found.
The study, published in the journal Nature Machine Intelligence, could help the team to learn more about how the way the human brain is wired helps to support specific cognitive skills.
“The project unifies two vibrant and fast-paced scientific disciplines,” said Bratislav Misic, a researcher at The Neuro and the paper’s senior author.
“Neuroscience and AI share common roots, but have recently diverged. Using artificial networks will help us to understand how brain structure supports brain function. In turn, using empirical data to make neural networks will reveal design principles for building better AI.
“So, the two will help inform each other and enrich our understanding of the brain.”
Separate research from the University of Delaware in the US found that coating sensors and other bioelectronic devices in polymer could help them to integrate more easily into the human body.
The researchers were seeking a way to insert microelectrodes containing gold, stainless steel and silicon into human brain tissue without causing scarring, which also affects their electric signals and prevents them from operating properly when they created the coating.
Coating the components in the organic polymer, known as PEDOT, which was sold commercially as an anti-static coating for electronic displays, allowed the signals to flow freely, increasing signal quality and the length of battery life.
Dr David Martin, who led the study, was confident the research could contribute to the future merger of the human brain with AI, enhancing human capabilities with software and machine insight.
“The ability to do the polymerization in a controlled way inside a living organism would be fascinating,” he said.