English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Anomaly detection in Minimally-Invasive Surgery (MIS) traditionally requiresa human expert monitoring the procedure from a console. Data scarcity, on theother hand, hinders what would be a desirable migration towards autonomousrobotic-assisted surgical systems. Automated anomaly detection systems in thisarea typically rely on classical supervised learning. Anomalous events in asurgical setting, however, are rare, making it difficult to capture data totrain a detection model in a supervised fashion. In this work we thus proposean unsupervised approach to anomaly detection for robotic-assisted surgerybased on deep residual autoencoders. The idea is to make the autoencoder learnthe 'normal' distribution of the data and detect abnormal events deviating fromthis distribution by measuring the reconstruction error. The model is trainedand validated upon both the publicly available Cholec80 dataset, provided withextra annotation, and on a set of videos captured on procedures usingartificial anatomies ('phantoms') produced as part of the Smart AutonomousRobotic Assistant Surgeon (SARAS) project. The system achieves recall andprecision equal to 78.4%, 91.5%, respectively, on Cholec80 and of 95.6%, 88.1%on the SARAS phantom dataset. The end-to-end system was developed and deployedas part of the SARAS demonstration platform for real-time anomaly detectionwith a processing time of about 25 ms per frame.

Unsupervised anomaly detection for a Smart Autonomous Robotic Assistant Surgeon (SARAS)using a deep residual autoencoder