AI detects surgical site infections accurately from patient wound images.

An innovative AI system accurately detects surgical site infections (SSI) from patient-submitted postoperative wound images, streamlining clinician workload. Among 6,060 patients, the Vision Transformer model achieved a remarkable incision detection accuracy of 94% and an SSI detection accuracy of 73%. It demonstrated effectiveness across diverse racial subgroups, reinforcing its potential for equitable healthcare delivery. This automated approach not only enhances the speed of SSI identification but also significantly improves postoperative management.