IIT Guwahati’s algorithm to encode brain networks of healthy humans, patients with Parkinson’s

According to officials, the researchers adopted a network perspective, representing different brain regions as nodes and establishing connection values of the network based on regional grey matter volume.

Further, connection values for every node were weighted to capture the significance of each link by following a series of algorithmic steps, they said.

The obtained numerical representation (UBNIN) was observed to be distinct for each individual brain network, and also applicable to other neuroimaging brain modalities, they said.

This innovative research holds immense potential in the realm of brain printing and emerges as a promising biomarker with a numerical value for tracking mental illness progression over time, they said.

Parkinson’s disease, a neurodegenerative disorder, with clinical symptoms such as tremors, stiffness, and slow movement, worsens with age. However, neurodegeneration begins long before these symptoms appear, making early detection imperative for effective PD management.

“UBNIN is a special number representing unique characteristics of each human brain from a network perspective. Interestingly, we can also reverse engineer any human’s UBNIN value to reconstruct the original brain network,” said Cota Navin Gupta, Assistant Professor, Neural Engineering Lab, Department of Biosciences and Bioengineering, the Indian Institute of Technology (IIT), Guwahati.

“This UBNIN algorithm will enable us to identify and characterise (encode-decode) brain networks of every human being efficiently “Applying the UBNIN algorithm on longitudinal neuroimaging data (over time) holds promise for elucidating the dynamics of brain plasticity (changes in the human brain). This insight is crucial to understand how the human brain degenerates, and copes with damage due to underlying neurological diseases,” Gupta added.

The developed UBNIN algorithm makes MRI data interpretable and holds the potential to transform neurodegenerative disorder diagnosis and treatment. This may be used as a biomarker to complement other diagnostic tests recommended by neurologists.

“UBNIN’s applications range from brainprinting to optimising storage for structural MRI brain networks. This could open avenues for low-bandwidth, high-speed information transfer of human brain networks for Telemedicine and related applications. UBNIN’s adaptability could also be extended to other neuroimaging modalities like electroencephalogram (EEG), functional MRI (both resting and task-based), etc.

“It can also be applied to other neurological conditions like Schizophrenia, Alzheimer’s, Depression, etc. Furthermore, it may be implemented on various datasets such as protein, social and traffic networks, making it a versatile tool for understanding complex system dynamics. We are now looking into the possibilities of using UBNIN as a potential biomarker to distinguish healthy and Parkinson’s at the group level,” Gupta said. PTI GJS RHL