SurgeonAI

Landing
OpenSource
Launch???
Info
Info

Inspiration The realization that forty percent of surgical trainees lack sufficient practice due to a severe shortage of attending surgeons inspired the creation of a real-time, AI-driven virtual mentor.

What it does The platform uses a standard webcam to track a trainee's scalpel in real time, instantly providing AI-generated voice coaching if their approach angle deviates from the optimal safe zone.

How we built it We integrated a YOLOv8 and OpenCV computer vision pipeline on a Flask backend with a low-latency React interface and ElevenLabs text-to-speech to deliver instant geometric and auditory feedback.

Challenges we ran into Overcoming the massive bottleneck of processing heavy machine learning inferences, calculating spatial geometry, and streaming synchronous audio and visual feedback at thirty frames per second without latency was our biggest hurdle.

Accomplishments that we're proud of We successfully orchestrated a complex pipeline where machine learning, WebSockets, and audio generation perfectly synchronize to provide high-fidelity surgical tracking using only standard web cameras and accessible tools.

What we learned We learned how to optimize live video streams under heavy data loads and realized how accessible software paired with 3D-printed hardware can genuinely democratize surgical training globally.

What's next for SurgeonAI We plan to expand the AI protocol to support multi-step procedures and distribute the open-source hardware files to medical students in developing regions to gather real-world efficacy data.