A new artificial intelligence system successfully detects and localises early neoplasia in Barrett's esophagus by using convolutional neural networks

Abstract

Background and aims: Seattle protocol biopsies for Barrett's Esophagus (BE) surveillance are labour intensive with low compliance. Dysplasia detection rates vary, leading to missed lesions. This can potentially be offset with computer aided detection. We have developed convolutional neural networks (CNNs) to identify areas of dysplasia and where to target biopsy.Methods: 119 Videos were collected in high-definition white light and optical chromoendoscopy with i-scan (Pentax Hoya, Japan) imaging in patients with dysplastic and non-dysplastic BE (NDBE). We trained an indirectly supervised CNN to classify images as dysplastic/non-dysplastic using whole video annotations to minimise selection bias and maximise accuracy. The CNN was trained using 148,936 video frames (31 dysplastic patients, 31 NDBE, two normal esophagus), validated on 25,161 images from 11 patient videos and tested on 264 iscan-1 images from 28 dysplastic and 16 NDBE patients which included expert delineations. To localise targeted biopsies/delineations, a second directly supervised CNN was generated based on expert delineations of 94 dysplastic images from 30 patients. This was tested on 86 i-scan one images from 28 dysplastic patients. Findings:The indirectly supervised CNN achieved a per image sensitivity in the test set of 91%, specificity 79%, area under receiver operator curve of 93% to detect dysplasia. Per-lesion sensitivity was 100%. Mean assessment speed was 48 frames per second (fps). 97% of targeted biopsy predictions matched expert and histological assessment at 56 fps. The artificial intelligence system performed better than six endoscopists.This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

UEG week and UEG journal: A strong scientific liaison

“…Hussein et al evaluated the use of convolutional neural networks (CNNs) to assist the endoscopists in identifying areas of dysplasia (i.e. : target biopsy) 11 . Notably, the CNNs‐guided approach allowed identification of areas for biopsy with excellent diagnostic accuracy (higher than six expert endoscopists).…”

“…: target biopsy). 11 Notably, the CNNs‐guided approach allowed identification of areas for biopsy with excellent diagnostic accuracy (higher than six expert endoscopists). Pending further validation, these results support the application of artificial intelligence in daily life clinical practice.…”

Automatic esophagus Z‐line delineation in endoscopic images using a new boundary linking method

“…Hussein et al. developed a neural network model to annotate early neoplasia in Barrett's esophagus [16]. As the first step the high‐definition endoscopic images are classified as dysplastic/non‐dysplastic using a supervised neural network.…”

“…Various CAD methods have been presented using white light endoscopic images, which can be categorized into classification, Region of interest (ROI) detection, and annotation. The classification category methods aim to classify between the images with normal and abnormal regions and detect the type of the disorder [11][12][13][14][15][16]. These works have employed either deep learning tools [11][12][13][14][15][16][17][18][19][20][21][22][23] or conventional image processing techniques [22][23][24][25][26][27].…”

Artificial intelligence in Barrett's oesophagus and the need for shared and combined data

“…In the latest issue of UEGJ, a new study adds further evidence concerning the potential of AI in BE surveillance. Hussein et al 10 . investigate different technical approaches and application scenarios.…”