P178 A novel CADx algorithm for characterisation and sizing of colorectal polyps meets PIVI 1 and PIVI 2 threshold

Introduction

There is increasing evidence for the use of ‘resect and discard’ for colorectal polyps, however this has not been widely adopted due to the challenges of accurate optical diagnosis and sizing of colorectal polyps outside of expert centres. Computer aided diagnosis (CADx) has shown promising potential for improving optical diagnosis. However, there is still no data on real-time sizing and characterisation together which is what will be required for the introduction of this strategy and meeting PIVI threshold.

Methods

A dedicated colon CADx system was developed to categorise polyps by histology (neoplastic or non-neoplastic) and by size (diminutive or non-diminutive). In the first phase of this study, the system was tested on a customised video based platform using prospectively collected endoscopy videos. Phase 2 tested real-time performance of CADx during colonoscopy. Ground truth in both phases was expert endoscopist assessment and histology.

Results

In video based analysis, 292 polyp videos (202 neoplastic and 90 non-neoplastic) were assessed using the CADx device. In real-time analysis, 116 (66 neoplastic and 50 non-neoplastic) polyps in 36 patients were assessed with CADx during real-time colonoscopy. Results for CADx in video based and real-time use are shown in table 1. Video based AI sizing demonstrated sensitivity, specificity, accuracy and NPV for non-diminutive size of 93.40%, 79.03%, 93.40% and 95.45% respectively. Real-time concordance for sizing between ground truth and AI as diminutive or non-diminutive was 84.61%. Subgroup analysis was carried out for video based and real-time CADx for performance on diminutive polyps. The NPV for diminutive rectosigmoid polyps for adenomatous diagnosis was 93.3% in real-time. There was 97.2% concordance between surveillance intervals (ASGE and BSG guidelines) based on histology versus AI diagnosis for diminutive polyps plus histology for non-diminutive.

Conclusions

This is the first study demonstrating the outcome of AI based sizing and characterisation of polyps in real time. The system meets PIVI 1 and PIVI 2 threshold suggesting we are close to being able to implement ‘resect and discard’ and ‘diagnose and leave strategies’. Further validation from a multicentre prospective study will help guide this.