Novel application of artificial intelligence to automate measurement of colonoscopy inspection time

Introduction

A randomised controlled trial recently demonstrated that adenoma detection rate (ADR) increases as colonoscopy inspection time (CIT) rises from 6 to 13 minutes ¹. CIT was defined as withdrawal from caecum to the anal verge with the timer manually stopped during cleaning (washing or suctioning) and polypectomy events. However, CIT measurement cannot be implemented in clinical practice without dedicated staff to pause a timer during these events. Previous work demonstrated artificial intelligences (AI) ability to measure withdrawal time (WT) by detection of caecal landmarks and the anal verge. We aimed to develop convolutional neural networks (CNN) to detect cleaning and polypectomy phases during withdrawal to facilitate automated measurement of CIT.

Methods

Videos of withdrawal were annotated with labels of ‘cleaning’ (suctioning and washing), ‘polypectomy’ (injection, polypectomy and inspecting resection margins) and ‘inspection time’. Using 98 annotated videos, two ResNet-101 CNNs were developed to detect cleaning (213,936 frames) and polypectomy (56,273 frames) events. The CNNs was then evaluated with 40 consecutive colonoscopy videos from 5 expert (ADR>45%) and 40 videos from 6 non-expert (ADR<30%) endoscopists, with the annotations referenced as the ground truth.

Results

For the test set of 80 procedures, the median WT documented in the endoscopy reporting system (EPIC) was 15.0 minutes (IQR 10.0–23.0), annotated ground truth median CIT was 7.1 minutes (IQR 5.3–9.6), and AI predicted median CIT was 6.9 minutes (IQR 5.1–9.0) with a correlation coefficient (r) of 98.2%. For expert procedures (n=40), these were 16.5 minutes (IQR 11.0–25.3), 7.1 minutes (IQR 5.1–8.6) and 6.9 minutes (IQR 4.9–8.8), respectively. Amongst non-expert procedures (n=40), these were 14.0 minutes (IQR 10.0–22.0), 7.0 minutes (IQR 5.8–10.7) and 7.0 minutes (IQR 5.3–10.2), respectively. The AI system correctly categorised 95% (76/80) of procedures as less or more than 6 minutes CIT.

Conclusions

We have demonstrated the feasibility of AI to differentiate the phases of withdrawal to automate measurement of CIT and the considerable difference in time between CIT and the WT documented in endoscopy reporting system.