Review highlights AI’s promise for improving preoperative, intraoperative, postoperative care

Editor's Note

Although artificial intelligence (AI) applications in surgery “remain relatively nascent,” the technology has potential to significantly impact all phases of surgical care, according to a review article published May 13 in Nature Medicine. “The emergence of foundation model architectures, wearable technologies, and improving surgical data infrastructures is enabling rapid advances in AI interventions and utility,” the authors write. “Maturing AI methods hold the potential to improve patient outcomes, facilitate surgical education and optimize surgical care.”

In the preoperative space, AI presents opportunities in diagnostics, risk assessment, patient selection, operative planning, and patient counseling. Examples cited in the article include improving breast cancer diagnostics and detecting pancreatic cancer on CT scans. AI also could help characterize patients' physiological and psychosocial factors, as well as drive patient-facing tools for informed consent, education, and postoperative follow-up.

During the intraoperative period, computer vision could help with analyzing anatomy, identifying pathology, and predicting procedural next steps, laying the groundwork for future autonomous systems. Authors paint a picture of an operating room that leverages real-time access to patient-specific information and other data for to improve teamwork, guide decision-making, improve surgical skill assessments and training, and help drive surgical robots and other automation.   

During the postoperative period, AI-driven hospital-at-home services could improve equity and access to healthcare while optimizing patient recovery and reducing hospital burdens. Wearable devices present the possibility of continuous patient monitoring for more personalized discharge planning and rehabilitation goals. AI-based technologies also show promise in early detection and intervention of postoperative complications. Examples cited in the article include anastomotic leaks and pancreatic fistulas.  

Overall, AI in surgery holds great promise, but significant challenges remain, the authors conclude. Noting that only six randomized trials of AI exist in surgery, they emphasize the need for further study, exploration of novel use cases, development of digital infrastructure, and more to realize the technology’s full potential.