Event Recognition in Laparoscopic Gynecology Videos with Hybrid Transformers

TL;DR

This paper introduces a hybrid transformer-based approach for recognizing key events in laparoscopic gynecology videos, utilizing a new annotated dataset and a frame sampling strategy to improve accuracy amidst challenging surgical conditions.

Contribution

The paper presents a novel hybrid transformer architecture and a specialized training-inference framework for event recognition in laparoscopic videos, along with a new dataset and sampling strategy.

Findings

Hybrid transformer architecture outperforms CNN-RNN models in accuracy.

The dataset includes annotations for intra-operative challenges and complications.

Frame sampling strategy enhances temporal resolution and robustness.

Abstract

Analyzing laparoscopic surgery videos presents a complex and multifaceted challenge, with applications including surgical training, intra-operative surgical complication prediction, and post-operative surgical assessment. Identifying crucial events within these videos is a significant prerequisite in a majority of these applications. In this paper, we introduce a comprehensive dataset tailored for relevant event recognition in laparoscopic gynecology videos. Our dataset includes annotations for critical events associated with major intra-operative challenges and post-operative complications. To validate the precision of our annotations, we assess event recognition performance using several CNN-RNN architectures. Furthermore, we introduce and evaluate a hybrid transformer architecture coupled with a customized training-inference framework to recognize four specific events in laparoscopic surgery videos. Leveraging the Transformer networks, our proposed architecture harnesses inter-frame dependencies to counteract the adverse effects of relevant content occlusion, motion blur, and surgical scene variation, thus significantly enhancing event recognition accuracy. Moreover, we present a frame sampling strategy designed to manage variations in surgical scenes and the surgeons' skill level, resulting in event recognition with high temporal resolution. We empirically demonstrate the superiority of our proposed methodology in event recognition compared to conventional CNN-RNN architectures through a series of extensive experiments.