Performance of a Deep Learning System for Automatic Diagnosis of Protruding Lesions in Colon Capsule Endoscopy

Abstract

Background: Colon capsule endoscopy (CCE) is an alternative for patients unwilling or with contraindications for conventional colonoscopy. Colorectal cancer screening may benefit greatly from widespread acceptance of a non-invasive tool such as CCE. However, reviewing CCE exams is a time-consuming process, with risk of overlooking important lesions. We aimed to develop an artificial intelligence (AI) algorithm using a convolutional neural network (CNN) architecture for automatic detection of colonic protruding lesions in CCE images. An anonymized database of CCE images collected from a total of 124 patients was used. This database included images of patients with colonic protruding lesions or patients with normal colonic mucosa or with other pathologic findings. A total of 5715 images were extracted for CNN development. Two image datasets were created and used for training and validation of the CNN. The AUROC for detection of protruding lesions was 0.99. The sensitivity, specificity, PPV and NPV were 90.0%, 99.1%, 98.6% and 93.2%, respectively. The overall accuracy of the network was 95.3%. The developed deep learning algorithm accurately detected protruding lesions in CCE images. The introduction of AI technology to CCE may increase its diagnostic accuracy and acceptance for screening of colorectal neoplasia.

Capsule endoscopy and panendoscopy: A journey to the future of gastrointestinal endoscopy

“…To date, this has been mainly tested for the small bowel[ 69 , 70 ], while it remains scarcely explored in the case of double-headed capsules. A few studies of CNN development for CC endoscopy revealed a very high accuracy for detection of colorectal neoplasia or protruding lesions[ 71 - 73 ]. Mascarenhas et al [ 74 ] recently developed a CNN for automatic detection of colonic blood in CC endoscopy, enabling the differentiation between normal mucosa, blood or hematic residues and pleomorphic mucosal lesions, namely ulcers and erosions, protruding lesions and vascular lesions.…”

Artificial intelligence in capsule endoscopy: development status and future expectations

“…Deep learning was integrated into the field with the study by Yuan and Meng in 2017 [62] , where they utilised a stacked sparse autoencoder method to categorise images into polyps, bubbles, turbid images, and clear images with an overall accuracy of 98.00%. Since then, 12 deep learning applications were used for polyp and tumour detection [63][64][65][66][67][68][69][70][71][72][73][74] . More recently, a study by Lafraxo et al in 2023 proposed an innovative model using CNN (Resnet50), where they achieved an accuracy of 99.16% on the MICCAI 2017 WCE dataset [73] .…”

Clinicians’ Guide to Artificial Intelligence in Colon Capsule Endoscopy—Technology Made Simple

“…AI in colon capsule endoscopy is a new field of interest. Recently, Afonso et al [ 26 ] analysed 24 CCE exams (PillCam ® COLON 2) performed at a single centre between 2010 and 2020. From these video recordings, 3635 frames of the colonic mucosa were extracted, 770 containing colonic ulcers or erosions and 2865 showing normal colonic mucosa.…”

“… Study No. of Images Colonic Lesion Normal Colonic Mucosa Sensitivity Specificity Accuracy of the Network AUROC for Detection of Protruding Lesion Afonso [ 26 ] 3635 770 2865 90.3% 98.8% 97.0% 0.99 Saraiva [ 2 ] 5715 2410 3305 90.0% 99.1% 95.3% 0.99 Atsuo Yamada [ 27 ] 4784 1850 2934 79.0% 87.0% 0.902 Hiroaki Saito [ 28 ] 17,507 7507 10,000 90.0% 79.0% 0.911 Nadimi, E.S [ 14 ] 1695 4800 6500 …”