Sichuan University's West China Second Hospital demonstrated an AI-powered clinical skills examiner at a primary-care pediatric training session in Nanchong on November 13, Chengdu-based outlet Chuanguan News reported on November 14. The platform, unveiled as an AI teaching aid for clinical skills, has already been piloted at 12 primary hospitals across 11 provinces, training more than 1,000 grassroots doctors and supporting over 100 physicians on rotations.
The examiner is built from a sensor array: pressure sensors log the depth and rate of chest compressions, cameras watch positioning, garment access and clearing of foreign objects, and audio sensors time the rescue call and retrieval of an AED, after which the system scores performance against expert benchmarks. Liu Hanmin, the hospital's president, said emergency skills "go dormant" without practice — the platform keeps reminding doctors to revisit them, flags compressions that are too fast, slow, deep or shallow, and by the time a trainee scores in the 80-to-90-point range, the muscle memory has formed.
The examiner grows out of the hospital's "Liu Shou Ren" program, which fields 7 instructors and more than 200 volunteers and has been deploying to hospitals in Tibet and Ya'an since 2024. The platform will cover 24 skills in line with the national physician licensing exam syllabus. Basic functions are free, while advanced training and clinical skills centers are paid services whose revenue flows back to remote regions; project lead Yang Wenhao said the team is entering innovation competitions and lining up financing.
The old failure mode of medical outreach is expertise leaving with the experts. Turning assessment and refresher training into sensor-backed infrastructure is an attempt to freeze expert standards into reusable form. Province-wide rollout in Sichuan and full coverage in Tibet are the program's next targets — and the two checkpoints where its sustainability will actually be tested.