Post-surgical Endometriosis Segmentation in Laparoscopic Videos

2025-10-14

AI Summary

Endometriosis exhibits heterogeneous clinical phenotypes, and deeply pigmented lesions are challenging to identify in laparoscopic video, leading to frequent underdiagnosis. To address this, we propose the first dedicated semantic segmentation system for dark endometriotic lesions, integrating deep learning with temporal frame-sequence analysis to achieve pixel-level lesion delineation. Our method introduces two key innovations: (1) multi-color coverage annotation to enhance lesion boundary discrimination, and (2) a structured detection summary module that overlays segmented lesion regions onto raw video frames while generating key-frame localization and quantitative lesion statistics. Evaluated on real-world laparoscopic video datasets, the system significantly improves lesion detection accuracy and video review efficiency. It supports gynecologists in precise intraoperative lesion localization, real-time surgical decision-making, and postoperative case review—thereby advancing intelligent assistance in minimally invasive surgery.

Technology Category

• Computer Vision: Segmentation
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Multi-instance/Multi-view Learning

Application Category

• Security and Privacy: Data transparency and provenance
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Endometriosis is a common women's condition exhibiting a manifold visual appearance in various body-internal locations. Having such properties makes its identification very difficult and error-prone, at least for laymen and non-specialized medical practitioners. In an attempt to provide assistance to gynecologic physicians treating endometriosis, this demo paper describes a system that is trained to segment one frequently occurring visual appearance of endometriosis, namely dark endometrial implants. The system is capable of analyzing laparoscopic surgery videos, annotating identified implant regions with multi-colored overlays and displaying a detection summary for improved video browsing.

Problem

• Segmenting dark endometrial implants in laparoscopic videos
• Assisting gynecologists with endometriosis identification
• Automating surgical video annotation for improved diagnosis

Innovation

• Segments dark endometrial implants in videos
• Analyzes laparoscopic surgery videos automatically
• Annotates detected regions with multi-colored overlays