Learning Function from Neuromorphic Network Structure
The neural circuit connection design in the brain is quite complex. New tech development allows the reconstruction of human brain networks on the macroscale. The functional & computation specialization developed from the network design - connectomes that are densely interconnected core are still unknown.
The connectomes are reconstructed leveraging in a Vivo diffusion-weighted imaging & leverage reservoir computing to execute the brain network as artificial neural networks. These networks can be trained to learn encoding tasks. These architectures perform top-notch when they portray crucial dynamics.
Its performance is directed by network topology & the modular architecture is relevant, observing a vital connection between network dynamics & structure. This innovative tech works great with new-flanged opportunities to identify how the network organization optimizes cognitive capacity.
AI Network based on Biological Brain can Perform Cognitive Tasks
When the MRI data is monitored from a large Open Science repository, scientists develop a brain connectivity pattern to apply it on an Artificial Neural Network (ANN). The ANN is trained to execute a cognitive memory task & observed its operation for assignment completion.
Connectomics focuses on explaining the brain organization without diving into its computation & functional performance. The conventional ANNs have an arbitrary structure that does not consider the entire organization of the brain network. By combining this connectomics into the ANN architectural construction, scientists thought that digging deep into the brain's wiring supports specific cognitive skills.
Neuromorphic Neural Network
The scientists identified ANNs with biological brain connectivity - neuromorphic neural networks perform the memory tasks more professionally than other standard designs. Moreover, these network uses the same underlying design to support numerous learning potentials across various contexts.
A scientist at The Neuro & a senior author of the study - Bratislav Misic, says that the research unites 2 swift-paced scientific disciplines. Artificial intelligence & neuroscience share common ties but have diverged recently. Using AI networks, one can get an insight into how the brain structure supports its function.
The Neuro, a World Leading Destination for Brain Research
It was found by a neurosurgeon - Dr. Wilder Penfield, in 1934. Since its founding, the firm - The Neuro has extended to be the most extensive specialized neuroscience research & clinical center in Canada & one of the biggest across the globe. The firm's research combination, training sessions, & patient care make it unique to significantly affect nervous system disorder treatments.
The institute became the first to fully accept the Open Science philosophy in 2016 across the world, developing the Tanenbaum Open Science Institute.
Image Source: Unsplash.com