Applications of Deep Learning Models in Laparoscopy for Gynecology

Background and Objectives: The use of Artificial Intelligence (AI) in the medical field is rapidly expanding. This review aims to explore and summarize all published research on the development and validation of deep learning (DL) models in gynecologic laparoscopic surgeries. Materials and Methods: MEDLINE, IEEE Xplore, and Google scholar were searched for eligible studies published between January 2000 and May 2025. Selected studies developed a DL model using datasets derived from gynecologic laparoscopic procedures. The exclusion criteria included non-gynecologic datasets, non-laparoscopic datasets, non-Convolutional Neural Network (CNN) models, and non-English publications. Results: A total of 16 out of 621 studies met our inclusion criteria. The findings were categorized into four main application areas: (i) anatomy classification (n = 6), (ii) anatomy segmentation (n = 5), (iii) surgical instrument classification and segmentation (n = 5), and (iv) surgical action recognition (n = 5). Conclusions: This review emphasizes the growing role of AI in gynecologic laparoscopy, improving anatomy recognition, instrument tracking, and surgical action analysis. As datasets grow and computational capabilities advance, these technologies are poised to improve intraoperative guidance and standardize surgical training.