Benchmarking and Enhancing Surgical Phase Recognition Models for Robotic-Assisted Esophagectomy

TL;DR

This paper introduces a new dataset and a novel deep learning model with hierarchical attention for improved surgical phase recognition in complex robotic-assisted esophagectomy procedures, aiming to support surgeons intraoperatively.

Contribution

The paper presents a new dataset and a novel encoder-decoder deep learning model with hierarchical attention for better phase recognition in RAMIE surgeries.

Findings

The new model outperforms existing models in phase recognition accuracy.

The dataset enables more comprehensive evaluation of recognition models.

Hierarchical attention improves temporal dynamics modeling.

Abstract

Robotic-assisted minimally invasive esophagectomy (RAMIE) is a recognized treatment for esophageal cancer, offering better patient outcomes compared to open surgery and traditional minimally invasive surgery. RAMIE is highly complex, spanning multiple anatomical areas and involving repetitive phases and non-sequential phase transitions. Our goal is to leverage deep learning for surgical phase recognition in RAMIE to provide intraoperative support to surgeons. To achieve this, we have developed a new surgical phase recognition dataset comprising 27 videos. Using this dataset, we conducted a comparative analysis of state-of-the-art surgical phase recognition models. To more effectively capture the temporal dynamics of this complex procedure, we developed a novel deep learning model featuring an encoder-decoder structure with causal hierarchical attention, which demonstrates superior performance compared to existing models.