Artificial intelligence and polyp detection in colonoscopy: Use of a single neural network to achieve rapid polyp localization for clinical use

Abstract

Background and Aim Artificial intelligence has been extensively studied to assist clinicians in polyp detection, but such systems usually require expansive processing power, making them prohibitively expensive and hindering wide adaption. The current study used a fast object detection algorithm, known as the YOLOv3 algorithm, to achieve real‐time polyp detection on a laptop. In addition, we evaluated and classified the causes of false detections to further improve accuracy. Methods The YOLOv3 algorithm was trained and validated with 6038 and 2571 polyp images, respectively. Videos from live colonoscopies in a tertiary center and those obtained from public databases were used for the training and validation sets. The algorithm was tested on 10 unseen videos from the CVC‐Video ClinicDB dataset. Only bounding boxes with an intersection over union area of > 0.3 were considered positive predictions. Results Polyp detection rate in our study was 100%, with the algorithm able to detect every polyp in each video. Sensitivity, specificity, and F1 score were 74.1%, 85.1%, and 83.3, respectively. The algorithm achieved a speed of 61.2 frames per second (fps) on a desktop RTX2070 GPU and 27.2 fps on a laptop GTX2060 GPU. Nearly a quarter of false negatives happened when the polyps were at the corner of an image. Image blurriness accounted for approximately 3% and 9% of false positive and false negative detections, respectively. Conclusion The YOLOv3 algorithm can achieve real‐time poly detection with high accuracy and speed on a desktop GPU, making it low cost and accessible to most endoscopy centers worldwide.

Application of artificial intelligence in diagnosis and treatment of colorectal cancer: A novel Prospect

“…It could accomplish an excellent effect in a short time. It was economical and affordable, making it suited for large-scale promotions in underdeveloped areas ( 88 ). Some adenomas and polyps detected by the real-time CADe system are small and low risk.…”

Is this the end of colonoscopy screening for colorectal cancer? An Asia‐Pacific perspective

“…The software, hardware, and interface platforms are already available. A study showed that a commercially available desktop graphics processing unit can already run some of these algorithms, making it attractive for low‐middle‐income countries 28 . This may allow for technology leapfrogging if colonoscopy‐based AI‐enhanced screening programs become more mainstream.…”

Artificial intelligence-assisted colonoscopy: a narrative review of current data and clinical applications

“…This could be due to being hidden behind mucosal folds or concealed by poor bowel preparation. In a study that evaluated false detections with CADe in colonoscopy, (38) blurry images were found to result in distorted polyp texture and were one of the reasons for false negative detections. These false negative detections also occurred when polyps approached the corner of a frame just before appearing or disappearing from the FOV.…”