Unsupervised Anomaly Detection for a Smart Autonomous Robotic Assistant Surgeon (SARAS) Using a Deep Residual Autoencoder

Dinesh Jackson Samuel,Fabio Cuzzolin

doi:10.1109/lra.2021.3097244

Dinesh Jackson Samuel, Fabio Cuzzolin

Open Access

https://doi.org/10.1109/lra.2021.3097244

Copy DOI

Abstract

Anomaly detection in Minimally-Invasive Surgery (MIS) traditionally requires a human expert monitoring the procedure from a console, whereas automated anomaly detection systems in this area typically rely on classical supervised learning. Anomalous surgical events, however, are rare, making it difficult to capture data to train a model in a supervised fashion. In this work we propose an unsupervised approach to anomaly detection for robotic MIS based on deep residual autoencoders. The idea is to make the autoencoder learn the `normal' distribution of the data and detect abnormal events deviating from this distribution by measuring a reconstruction error. The model is trained and validated upon both the publicly available Cholec80 dataset and a set of videos captured on procedures using artificial anatomies (`phantoms') as part of the Smart Autonomous Robotic Assistant Surgeon (SARAS) project. The system achieves recall and precision equal to 78.4%, 91.5%, respectively, on Cholec80 and of 95.6%, 88.1% on the SARAS phantom dataset. The system was developed and deployed as part of the SARAS platform for real-time anomaly detection with a processing time of 25 ms per frame.

Highlights

I N RECENT years, Minimally-Invasive Surgery (MIS) has attracted a great deal of interest, as it only requires small incisions (5-30 mm) to provide the endoscope and other instruments access to the surgical cavity, rather than the vast ones demanded by traditional surgery
Performance was evaluated using both anomalous and nonanomalous frames generated by the Smart Autonomous Robotic Assistant Surgeon (SARAS) demonstration platform and an existing dataset in the surgery domain, Cholec80, adapted for anomaly detection
The videos are captured at 25 frames per second with a resolution of 854 × 480 pixels, and contain both anomalous and nonanomalous frames. As it was designed for phase recognition and tool detection rather than anomaly detection, in Cholec80 each frame is labelled with the phase of the procedure and the presence of tools

Summary

Introduction

I N RECENT years, Minimally-Invasive Surgery (MIS) has attracted a great deal of interest, as it only requires small incisions (5-30 mm) to provide the endoscope and other instruments access to the surgical cavity, rather than the vast ones (approximately 300 mm) demanded by traditional surgery. Robotic MIS (R-MIS) technology is adaptive, precise and accurate, most R-MIS systems are not designed to replace the main surgeon conducting the procedure but to increase the safety and effectiveness of surgeries. One such kind of human-machine interactive robotic system, named ‘da Vinci,’ has been developed by Intuitive Surgical to perform precise and complex surgeries through small incisions. Date of publication July 14, 2021; date of current version July 29, 2021. This letter was recommended for publication by Associate Editor E. Valdastri upon evaluation of the reviewers’ comments.

Methods

Results

Conclusion