Automatic Endoscopic Ultrasound Station Recognition with Limited Data

TL;DR

This paper presents a deep learning-based AI tool for real-time recognition of EUS stations during pancreatic cancer diagnosis, using minimal labeled data and providing interpretable visualizations to assist clinicians.

Contribution

It introduces a user-friendly web app for efficient annotation and demonstrates high accuracy with limited data, advancing AI-assisted training in endoscopic ultrasound procedures.

Findings

Achieved 89% balanced accuracy with only 43 procedures.

Developed an open-source annotation tool for minimal clinician effort.

Provided Grad-CAM visualizations for interpretability.

Abstract

Pancreatic cancer is a lethal form of cancer that significantly contributes to cancer-related deaths worldwide. Early detection is essential to improve patient prognosis and survival rates. Despite advances in medical imaging techniques, pancreatic cancer remains a challenging disease to detect. Endoscopic ultrasound (EUS) is the most effective diagnostic tool for detecting pancreatic cancer. However, it requires expert interpretation of complex ultrasound images to complete a reliable patient scan. To obtain complete imaging of the pancreas, practitioners must learn to guide the endoscope into multiple "EUS stations" (anatomical locations), which provide different views of the pancreas. This is a difficult skill to learn, involving over 225 proctored procedures with the support of an experienced doctor. We build an AI-assisted tool that utilizes deep learning techniques to identify these stations of the stomach in real time during EUS procedures. This computer-assisted diagnostic (CAD) will help train doctors more efficiently. Historically, the challenge faced in developing such a tool has been the amount of retrospective labeling required by trained clinicians. To solve this, we developed an open-source user-friendly labeling web app that streamlines the process of annotating stations during the EUS procedure with minimal effort from the clinicians. Our research shows that employing only 43 procedures with no hyperparameter fine-tuning obtained a balanced accuracy of 89%, comparable to the current state of the art. In addition, we employ Grad-CAM, a visualization technology that provides clinicians with interpretable and explainable visualizations.