Scientists Create One of the Most Detailed Digital Models of the Mouse Cortex Using Japan’s Fugaku Supercomputer

TechMintOra
By -
0

Scientists Build Ultra-Detailed Digital Mouse Cortex Using Japan’s Fugaku Supercomputer 


Researchers have taken a major step forward in computational neuroscience by creating one of the most detailed digital reconstructions of the mouse cortex ever produced. Using Japan’s world-class Fugaku supercomputer, the team built a virtual brain model containing around 10 million neurons and 26 billion synapses—a scale that captures both the structure and intricate behaviour of real cortical circuits.
The project was led by scientists from the Allen Institute in the U.S. and the University of Electro-Communications in Tokyo, combining cutting-edge biology with high-performance computing.


A Supercomputer-Powered Look Inside the Brain

To build the model, researchers integrated extensive biological datasets from the Allen Institute, including cell types, connectivity patterns, and anatomical structure. These inputs were processed using the institute’s Brain Modelling Toolkit, then executed on Fugaku’s massive computing infrastructure.

Each of the roughly 10 million neurons inside the simulation is represented as a branching, multi-compartment structure, allowing the model to capture electrical behaviours at the sub-cellular level.
When running at full capacity, Fugaku simulated the entire mouse cortex at a rate of 32 seconds of computation for every 1 second of real neural activity—a remarkable feat for a model of this complexity.


What This Digital Cortex Could Reveal

Scientists believe the simulated cortex could become a powerful tool for understanding brain disorders and testing new treatments. Because the model operates as a full network, researchers can observe how conditions like Alzheimer’s disease, epilepsy, and other neurological disorders spread through the brain—or how therapeutic interventions might stop them.

This work complements other global efforts to build detailed digital brain systems.

  • At Stanford, researchers have created an AI-based “digital twin” of the visual cortex capable of predicting how thousands of neurons react to new visual stimuli.
  • Meanwhile, teams at EPFL have generated full-brain synthetic wiring diagrams (digital connectomes) that closely match laboratory-measured data.

Together, these projects highlight how computational neuroscience is entering a new era—one where entire brain regions can be studied, manipulated, and understood without ever touching a living animal.


#Neuroscience #Supercomputing #Fugaku #AllenInstitute #BrainSimulation #AIResearch #DigitalTwin #ScienceNews #EPFL #StanfordResearch #MouseCortex


Tags:

Post a Comment

0 Comments

Post a Comment (0)

#buttons=(Ok, Go it!) #days=(20)

Our website uses cookies to enhance your experience. Check Now
Ok, Go it!