AI and optogenetics enable precision diagnosis of Parkinson’s

Parkinson’s disease is known for symptoms such as tremors, rigidity, bradykinesia, and postural instability. Traditional diagnostic methods often fail to capture subtle early-stage changes. The KAIST team introduced AI-powered 3D pose estimation, analyzing more than 340 behavioral features, including gait and limb movements. These were condensed into a single metric: the AI-predicted Parkinson’s disease score (APS).

• APS detected differences only two weeks after induction of Parkinson’s in mice.
• It outperformed conventional motor function tests in sensitivity.
• Key indicators included stride changes, asymmetrical limb use, chest tremors, and posture shifts.

To ensure APS was specific to Parkinson’s and not general motor decline, the team also tested mice with Amyotrophic Lateral Sclerosis (ALS). Despite showing motor impairments, ALS mice did not present high APS scores. This confirmed that APS reflects distinct Parkinson’s-related changes rather than generic movement problems.

For treatment evaluation, the researchers applied optoRET, an optogenetics technology that uses light to precisely regulate neurotrophic signals. In Parkinson’s mouse models, this resulted in:

• Improved gait and smoother limb movement
• Reduced tremors
• A protective effect on dopamine-producing neurons when applied on alternate days

Professor Won Do Heo of KAIST highlighted: “This is the first time in the world that a preclinical framework has been implemented that connects early diagnosis, treatment evaluation, and mechanism verification of Parkinson's disease by combining AI-based behavioral analysis with optogenetics.”

The findings provide a framework for personalized medicine, linking diagnosis with treatment evaluation and mechanistic insights.

MEDICA-tradefair.com; Source: KAIST (Korea Advanced Institute of Science and Technology)