Evaluating AIDriven 3D Surgical Models for Laparoscopic Strategy in Complex Endometriosis

Study Objective

Posterior compartment endometriosis carries up to 50% reoperation rates due to missed lesions. We evaluated whether AIgenerated 3D surgical models improve operative planning, anatomical validation, and intraoperative preparedness in complex, bowelinvolving disease.

Design

Prospective pilot study using structured surgeonreported outcomes and intraoperative validation across consecutive cases.

Setting

Multisite minimally invasive gynecology programs adopting the XENA surgical intelligence platform.

Patients or Participants

20 consecutive laparoscopic surgeries for suspected or confirmed moderatetosevere endometriosis.

Interventions

Preoperative review of patientspecific, AIgenerated 3D surgical models derived from standard pelvic MRI datasets.

The XENA platform transforms MRI data into operative planning maps tailored for surgical use—providing spatial insight to complement radiologic interpretation.

Measurements and Primary Results

Operating surgeons revised their laparoscopic approach in 9/20 cases (45%) after reviewing the 3D model, predominantly to adjust bowel or ureteric dissection strategy.

• Intraoperative validation confirmed 18/20 models (90%) accurately reflected surgical anatomy, reducing exploratory maneuvers.

• In 12/20 procedures (60%), surgeons noted that 3D models enabled more precise planning of dissection planes around critical pelvic structures.

Conclusion

AIdriven 3D surgical models demonstrate early clinical impact in highcomplexity endometriosis—particularly with bowel involvement—by enhancing planning precision and intraoperative orientation. These pilot data support integration of surgical intelligence platforms like XENA into preop workflows. A multicenter validation study is in protocol development.