Three-Dimensional Modeling of Deep Endometriosis From Pelvic MRI: Feasibility, Clinical Perception, and Preliminary
Claire Zhang1, Camille Martel1, Aline Host1
1Department of Gynecologic Surgery, University Hospitals of Strasbourg (Zhang, Martel, Host, Boisramé, Gabriele, Akladios and Lecointre), Strasbourg, France.
Objective:
To evaluate the feasibility of 3-dimensional (3D) modeling of deep endometriosis (DE) from pelvic MRI, to assess its perceived usefulness by gynecologic surgeons, and to explore its transposability in clinical practice.
Design:
Retrospective feasibility study with expert evaluation, national survey, and preliminary prospective cases.
Setting:
University hospital and national network of gynecologic surgeons.
Patients:
Consecutive women with surgically confirmed DE who underwent preoperative pelvic MRI with adequate image quality and an available radiology report between January 2010 and December 2022.
Interventions:
Segmentation of pelvic MRI sequences and creation of individualized 3D models.
Measurements And Main Results:
Among 147 patients, 41 MRI datasets allowed 3D reconstruction. The models included major pelvic organs and DE nodules. Two expert surgeons rated the added value as moderate to major in two-thirds of cases (κ = 0.41). In a national survey (n=31 respondents), the models were considered clear and precise (median 4/5). Compared with MRI alone, 3D modeling improved lesion visualization in 48% of cases, confirmed the surgical strategy in 67% and modified it in 8%. The perceived benefit was greater among less experienced surgeons. In 2 prospective cases, 3D models improved preoperative communication and patient understanding, although intraoperative use remained limited. Main barriers to wider use were cost and processing time.
Conclusion:
Large-scale 3D modeling of deep endometriosis from MRI is technically feasible and perceived as beneficial, especially in complex cases. Its main value lies in preoperative planning, surgical education, and patient counseling. Further prospective studies are required to validate its clinical impact.
