Accurate surgical phase recognition is essential for analyzing procedural workflows, supporting intraoperative decision-making, and enabling data-driven improvements in surgical education and performance evaluation. In this work, we present a comprehensive framework for phase recognition in pituitary tumor surgery (PTS) videos, combining self-supervised representation learning, robust temporal modeling, and scalable data annotation strategies. Our method achieves 90\% accuracy on a held-out test set, outperforming current state-of-the-art approaches and demonstrating strong generalization across variable surgical cases. A central contribution of this work is the integration of a collaborative online platform designed for surgeons to upload surgical videos, receive automated phase analysis, and contribute to a growing dataset. This platform not only facilitates large-scale data collection but also fosters knowledge sharing and continuous model improvement. To address the challenge of limited labeled data, we pretrain a ResNet-50 model using the self-supervised framework on 251 unlabeled PTS videos, enabling the extraction of high-quality feature representations. Fine-tuning is performed on a labeled dataset of 81 procedures using a modified training regime that incorporates focal loss, gradual layer unfreezing, and dynamic sampling to address class imbalance and procedural variability.
@article{arxiv.2603.24897,
title = {SurgPhase: Time efficient pituitary tumor surgery phase recognition via an interactive web platform},
author = {Yan Meng and Jack Cook and X. Y. Han and Kaan Duman and Shauna Otto and Dhiraj Pangal and Jonathan Chainey and Ruth Lau and Margaux Masson-Forsythe and Daniel A. Donoho and Danielle Levy and Gabriel Zada and Sébastien Froelich and Juan Fernandez-Miranda and Mike Chang},
journal= {arXiv preprint arXiv:2603.24897},
year = {2026}
}