RESEARCH PROJECT · ALON GORENSHTEIN, MD
INSPIRE: AI agents for nerve conduction studies and EMG
INSPIRE is a multi-agent AI framework developed by Alon Gorenshtein and colleagues to support interpretation and reporting of neuromuscular electrodiagnostic studies. It brings clinical context and electrophysiology information into a structured workflow for nerve conduction studies and EMG.
Research question
How can AI agents help electrophysiologists interpret electrodiagnostic findings and produce consistent reports?
Approach and evidence
The 2025 Clinical Neurophysiology paper describes a system that combines clinical information, electrophysiological data, and specialized agents to generate and evaluate interpretations. The development study compared the framework with a general-purpose language model.
A separate randomized trial evaluated physician reporting with and without AI assistance in 200 patients. The combined approach did not significantly improve report quality, and feedback identified usability and workload challenges. Together, the studies show why clinical workflow and human evaluation matter alongside model performance.
Authors and contributions
Alon Gorenshtein is first author of both the INSPIRE paper and the randomized trial. INSPIRE was developed with Moran Sorka, Mohamed Khateb, Dvir Aran, and Shahar Shelly.
Research context
INSPIRE is research in clinical decision support. The published studies should be consulted for their methods, results, and limitations; AI-generated reports require clinical oversight.
Publications and resources
Research summary updated October 27, 2026.