Artificial intelligence-assisted video analysis for evaluating ergonomic strain in robotic surgeons: a pilot study from the SAGES ergonomics task force.

BACKGROUND

Most surgeons experience work-related pain and musculoskeletal injuries (WRMSI). Though robotic surgery has been speculated to reduce the risk of WRMSI compared to open and laparoscopic approaches, the majority of robotic surgeons still experience neck, back, and upper extremity discomfort. Despite growing use of robot-assisted surgery (RAS), there is limited research on the ergonomic risks faced by robotic surgeons. This study aims to evaluate ergonomic strain during RAS.

METHODS

Artificial intelligence (AI)-assisted video analysis was used on a convenience sample of experienced faculty-level robotic surgeons to assess their positioning, practice, and pain. Participants completed 3 repetitions of a standardized simulated surgical task on the daVinci Xi robotic surgery console. Video recordings were analyzed using AI-assisted kinematic assessment software (TuMeKe Ergonomics, San Mateo CA). Validated subjective assessment tools and a general ergonomic practices questionnaire were calculated. Spearman's correlation coefficient evaluated the relationship between the time spent in medium-to-high-risk ergonomic positions and the corresponding reported discomfort.

RESULTS

The 5 participants had a mean age of 50.7 years and performed a mean of 3.5 robotic cases weekly. 60% were female. The average RULA score was 5.2 (SD 0.84), exceeding the "acceptable" threshold of 2. The trunk had the worst objective posture score (3/6), and participants spent 32% of the operative time in overall medium- to high-risk positions. There was significant correlation between reported discomfort and time spent in medium-to high-risk positions.

CONCLUSIONS

AI-assisted video analysis successfully identified high-risk ergonomic postures in robotic surgeons, correlating with subjective discomfort scores. This study highlights the need for ergonomic interventions in RAS and lays the groundwork for real-world intraoperative assessments across surgical modalities.