Beyond basic research, the model serves to study cognition, seizure propagation, brain wave synchronization, and inter-hemispheric interactions without invasive scans, providing a platform to test hypotheses about brain function and diseases.

The work was unveiled at the SC25 supercomputing conference and is published online, underscoring the potential to accelerate understanding of brain dynamics and disease mechanisms and enabling non-invasive in silico experimentation.

A collaborative team from the Allen Institute and the University of Electro-Communications in Japan built one of the most comprehensive virtual models of a mouse cortex to study brain function and disease.

The project integrates existing cellular data with new software to optimize processing and minimize unnecessary calculations, using the Fugaku supercomputer as its backbone while aiming to extend from regional to full-brain simulations.

Although smaller and simpler, the mouse brain shares key structural similarities with the human brain, making this model a valuable tool for neuroscience and studying diseases like Alzheimer’s.

The model enables controlled experiments not possible in living animals, potentially yielding deeper insights into neurological disorders and brain dynamics.

Anton Arkhipov from the Allen Institute called the achievement a technical milestone that demonstrates feasibility for much larger and more precise brain simulations in the future, including ambitions for whole-brain human models in virtual space.

The virtual brain simulates 9 million neurons, 26 billion synapses across 86 interconnected regions and can process quadrillions of calculations per second on the Fugaku supercomputer, offering unprecedented observation of brain activity.