Name
12387 - Manual Snapshots during Routine AI-Based Video Capture Can Increase Adoption of Safety Practices in Gynecologic Surgery
Presenting Author
Ishai Levin
Affiliation
Tel Aviv Sourasky Medical Center
Abstract
Study Objective:
Previous findings indicate that routine, automated capture and analysis of surgical video using computer vision-based surgical intelligence technology can increase adoption of safety-related practices during laparoscopic procedures. We propose that when practice adoption reaches a plateau, it can be further increased through deliberate capture of intraoperative snapshots. Here, we examined a quality intervention promoting intraoperative snapshots of both ureters after colpotomy and vault closure, aiming to improve safety in laparoscopic hysterectomy.
Design:
Prospective and retrospective analysis of videos from laparoscopic procedures.
Setting:
A high-volume gynecology department in a tertiary medical center.
Patients or Participants:
All total laparoscopic hysterectomies conducted between January and April 2024 were routinely recorded and stored by a surgical intelligence platform (Theator Inc.), which indicates for each video whether the ureters had been viewed.
Interventions:
Between January and March 2024, the surgical intelligence platform was used to routinely assess whether the ureters were viewed, in accordance with established departmental safety guidelines. In April 2024, a snapshot feature was added to the platform, enabling surgeons to manually capture still frames intraoperatively. A quality initiative was subsequently launched, incorporating snapshots of both ureters as a mandatory safety measure.
Measurements and Main Results:
In the 3 months before snapshots were introduced, the ureters were viewed in 78.9% of the the 19 laparoscopic hysterectomies conducted. Following the intervention, snapshots of ureter views were taken in all 8 of the procedures conducted, indicating 100% adoption of the safety measure.
Conclusion:
Serving as a promising pilot,
the preliminary findings collected since the recent implementation of the snapshot feature in a computer vision-based surgical intelligence platform indicate a significant increase in adoption of a key safety measure during laparoscopic hysterectomy. Continued monitoring will enable further validation and exploration of how intraoperative snapshots can contribute to surgical quality and safety.
Authors
Levin, I*1, Cohen, A1, Gil, Y2, Wolf, T3. 1Gynecology, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel; 2Tel Aviv Sourasky Medical Center, Tel Aviv, Israel; 3Theater Inc., Palo Alto, CA
Primary Category
New Instrumentation or Technology
Secondary Category
Basic Science/Education
Sponsorship Level
Virtual Poster
Logo
