Shin Kong Hospital Deploys AI to Flag Anesthesia Risks

Anesthesiologists must assess medical histories, medications and test results before surgery, but critical risk factors can be scattered across extensive clinical records. Shin Kong Hospital’s anesthesiology department has built a digital anesthesia-record system and a physiological-signal database, creating the data foundation for large language models to help clinicians identify preoperative risks and improve the speed and consistency of patient assessment.

The hospital said it has begun using AI and LLMs to extract potential anesthesia risk factors from patient records. Its next phase will apply AI to real-time physiological signals during operations, with the goal of detecting changes earlier and allowing clinicians to intervene sooner. Shin Kong Hospital did not disclose a launch date, the size of the underlying dataset or performance measures for the models.