A hospital analysis associated image-based operating-room AI and workflow changes with about 25 additional monthly cases across a 15-room surgical suite.
Houston Methodist reported a 7% increase in case volume across its 15-room cardiothoracic operating suite after introducing image-based AI to track operating-room events and support workflow decisions. The health system said the difference amounted to about 25 additional procedures a month over 10 months, without adding operating rooms.
The hospital used Apella’s platform to identify perioperative events and make workflow information available through dashboards, reports and scheduling tools. The system tracked activity that had often been recorded manually or retrospectively, giving operating-room leaders, anesthesia teams and other staff more visibility into room utilization, turnover and case progress. Surgeons could also receive automated notifications about readiness and procedures underway.
To estimate what might have happened without the technology, Houston Methodist’s team built a synthetic control: a comparison suite assembled from weighted data from other hospital operating rooms that had not implemented the platform. The comparison was calibrated to match the cardiothoracic suite’s case volume in the six months before launch. The hospital reported that the suite completed significantly more cases than this modeled comparison during the following 10 months.
The results do not mean that AI made individual procedures faster. According to the hospital’s account, total operative time was essentially unchanged and the case mix remained stable. The improvement was associated with lower turnover times and better use of existing operating-room capacity. The report also described a small reduction in median late-start minutes relative to the comparison, while noting that some secondary measures did not reach statistical significance.
The findings offer a practical example of AI being applied to hospital operations rather than diagnosis. They remain a health-system-reported evaluation, not a randomized trial, and the approach combined software with workflow improvements by staff. Other hospitals considering similar systems will need to assess privacy, implementation requirements and whether the results transfer to their own operating rooms and patient populations.