Scientists have developed a powerful new AI system called Cheese3D that can analyze facial expressions to understand what’s happening inside the brain. Created by researchers at Cold Spring Harbor Laboratory, the platform uses machine learning and advanced imaging to track even the tiniest facial muscle movements in mice—something that was previously impossible to measure accurately.
The system works by using six synchronized high-speed cameras to capture a mouse’s face from multiple angles. AI models then combine these recordings into a detailed 3D reconstruction, allowing researchers to measure subtle changes in facial muscle tone with extreme precision. This turns facial movement into a kind of “data language” that reflects internal brain activity.
One of the most striking findings is that the tool can predict the depth of anesthesia as accurately as traditional EEG brain scans, but without any invasive procedures. By simply observing facial expressions, researchers can determine whether an animal is awake, asleep, or somewhere in between—opening the door to safer and less intrusive monitoring methods.
Beyond anesthesia, the technology has far-reaching implications. Scientists believe it could help study brain development, emotional behavior, and neurological disorders like autism by linking facial expressions to underlying neural processes. In the long term, similar AI systems could be adapted for humans, offering a non-invasive way to diagnose conditions, monitor patients, and better understand how the brain controls behavior through facial expression.
