Artificial intelligence and capsule endoscopy: automatic detection of enteric protruding lesions using a convolutional neural network

Background and aims: capsule endoscopy (CE) revolutionized the study of the small intestine. Nevertheless, reviewing CE images is time-consuming and prone to error. Artificial intelligence algorithms, particularly convolutional neural networks (CNN), are expected to overcome these drawbacks. Protruding lesions of the small intestine exhibit enormous morphological diversity in CE images. This study aimed to develop a CNN-based algorithm for the automatic detection small bowel protruding lesions. Methods: a CNN was developed using a pool of CE images containing protruding lesions or normal mucosa from 1,229 patients. A training dataset was used for the development of the model. The performance of the network was evaluated using an independent dataset, by calculating its sensitivity, specificity, accuracy, positive and negative predictive values. Results: a total of 18,625 CE images (2,830 showing protruding lesions and 15,795 normal mucosa) were included. Training and validation datasets were built with an 80 %/20 % distribution, respectively. After optimizing the architecture of the network, our model automatically detected small-bowel protruding lesions with an accuracy of 92.5 %. CNN had a sensitivity and specificity of 96.8 % and 96.5 %, respectively. The CNN analyzed the validation dataset in 53 seconds, at a rate of approximately 70 frames per second. Conclusions: we developed an accurate CNN for the automatic detection of enteric protruding lesions with a wide range of morphologies. The development of these tools may enhance the diagnostic efficiency of CE (AU)

PRAPNet: A Parallel Residual Atrous Pyramid Network for Polyp Segmentation.

In a colonoscopy, accurate computer-aided polyp detection and segmentation can help endoscopists to remove abnormal tissue. This reduces the chance of polyps developing into cancer, which is of great importance. In this paper, we propose a neural network (parallel residual atrous pyramid network or PRAPNet) based on a parallel residual atrous pyramid module for the segmentation of intestinal polyp detection. We made full use of the global contextual information of the different regions by the proposed parallel residual atrous pyramid module. The experimental results showed that our proposed global prior module could effectively achieve better segmentation results in the intestinal polyp segmentation task compared with the previously published results. The mean intersection over union and dice coefficient of the model in the Kvasir-SEG dataset were 90.4% and 94.2%, respectively. The experimental results outperformed the scores achieved by the seven classical segmentation network models (U-Net, U-Net++, ResUNet++, praNet, CaraNet, SFFormer-L, TransFuse-L).

White-Light Endoscopic Colorectal Lesion Detection Based on Improved YOLOv5.

As an effective tool for colorectal lesion detection, it is still difficult to avoid the phenomenon of missed and false detection when using white-light endoscopy. In order to improve the lesion detection rate of colorectal cancer patients, this paper proposes a real-time lesion diagnosis model (YOLOv5x-CG) based on YOLOv5 improvement. In this diagnostic model, colorectal lesions were subdivided into three categories: micropolyps, adenomas, and cancer. In the course of convolutional network training, Mosaic data enhancement strategy was used to improve the detection rate of small target polyps. At the same time, coordinate attention (CA) mechanism was introduced to take into account channel and location information in the network, so as to realize the effective extraction of three kinds of pathological features. The Ghost module was also used to generate more feature maps through linear processing, which reduces the stress of learning model parameters and speeds up detection. The experimental results show that the lesion diagnosis model proposed in this paper has a more rapid and accurate lesion detection ability, and the AP value of polyps, adenomas, and cancer is 0.923, 0.955, and 0.87, and mAP@50 is 0.916.

Single Shot Multibox Detector Automatic Polyp Detection Network Based on Gastrointestinal Endoscopic Images.

PURPOSE: In order to resolve the situation of high missed diagnosis rate and high misdiagnosis rate of the pathological analysis of the gastrointestinal endoscopic images by experts, we propose an automatic polyp detection algorithm based on Single Shot Multibox Detector (SSD). METHOD: In the paper, SSD is based on VGG-16, the fully connected layer is changed to a convolutional layer, and four convolutional layers with successively decreasing scales are added as a new network structure. In order to verify the practicability, it is not only compared with manual polyp detection but also with Mask R-CNN. RESULTS: Multiple experimental results show that the mean Average Precision (mAP) of the SSD network is 95.74%, which is 12.4% higher than the manual detection and 5.7% higher than the Mask R-CNN. When detecting a single frame of image, the detection speed of SSD is 8.41 times that of manual detection. CONCLUSION: Based on the traditional pattern recognition algorithm and the target detection algorithm using deep learning, we select a variety of algorithms to identify and classify polyps to achieve efficient detection results. Our research demonstrates that deep learning has a lot of room for development in the field of gastrointestinal image recognition.

Two Types of Polyp Shape Observed in the Stomach of Patients with Peutz-Jeghers Syndrome.

The characteristics of gastric polyps in patients with Peutz-Jeghers (PJ) syndrome (PJS) have not been fully investigated. The objective of this study was to reveal the endoscopic and pathologic findings of gastric polyps in patients with PJS. We reviewed 11 patients with PJS treated at 6 institutions, and summarized the endo-scopic and pathologic features of their gastric polyps. The polyps were mainly classified into 2 types: (i) soli-tary or sporadic polyps > 5 mm, reddish in color with a sessile or semi-pedunculated morphology (n = 9); and (ii) multiple sessile polyps ≤ 5 mm with the same color tone as the peripheral mucosa (n = 9). Patients who underwent endoscopic mucosal resection for polyps > 5 mm were diagnosed with PJ polyps (n = 2), whereas those who underwent biopsy were diagnosed with hyperplastic polyps. Polyps ≤ 5 mm were pathologically diagnosed as fundic gland polyps or hyperplastic polyps. This study revealed that patients with PJS present with 2 types of polyps in the stomach. Endoscopic mucosal resection of polyps > 5 mm seems necessary for the pathologic diagnosis of PJ polyps.

The Workgroup Serrated Polyps and Polyposis (WASP) classification for optical diagnosis of colorectal diminutive polyps with iScan and the impact of the revised World Health Organization (WHO) criteria.

BACKGROUND AND AIMS: The Workgroup Serrated Polyps and Polyposis (WASP) developed criteria for optical diagnosis of colorectal polyps. The aims of this study were: (1) to improve optical diagnosis of diminutive colorectal polyps, especially SSLs, after training endoscopists in applying WASP criteria on videos of polyps obtained with iScan and (2) to evaluate if the WASP criteria are still useful when polyps are pathologically revised according to the World Health Organization (WHO) 2019 criteria. METHODS: Twenty-one endoscopists participated in a training session and predicted polyp histology on 30 videos of diminutive polyps, before and after training (T0 and T1 ). After three months, they scored another 30 videos (T2 ). Primary outcome was overall diagnostic accuracy (DA) at T0 , T1 and T2 . Polyps were histopathologically classified according to the WHO 2010 and 2019 criteria. RESULTS: Overall DA (both diminutive adenomas and SSLs) significantly improved from 0.58 (95% CI 0.55-0.62) at T0 to 0.63 (95% CI 0.60-0.66, p = 0.004) at T1 . For SSLs, DA did not change with 0.51 (95% CI 0.46-0.56) at T0 and 0.55 (95% CI 0.49-0.60, p = 0.119) at T1 . After three months, overall DA was 0.58 (95% CI 0.54-0.62, p = 0.787, relative to T0 ) while DA for SSLs was 0.48 (95% CI 0.42-0.55, p = 0.520) at T2 . After pathological revision according to the WHO 2019 criteria, DA of all polyps significantly changed at all time points. CONCLUSION: A training session in applying WASP criteria on endoscopic videos made with iScan did not improve endoscopists' long-term ability to optically diagnose diminutive polyps. The change of DA following polyp revision according to the revised WHO 2019 criteria suggests that the WASP classification may need revision.

Ileal Intussusception in an Adult Caused by a Locally Invasive Inflammatory Fibroid Polyp: A Case Report.

Inflammatory fibroid polyps (IFPs) are rare mesenchymal neoplasms affecting the gastrointestinal tract which are considered benign and noninvasive. We present a case of an invasive IFP in a 46-year-old woman who presented with signs of intestinal obstruction due to ileal intussusception. A segment of the small intestine was resected and subsequently intestinal continuity was restored. A polypoid lesion was found obstructing the lumen. Histopathology revealed a mesenchymal proliferation of spindle and stellate cells, without cytological atypia, arranged in a fibromyxoid stroma. The tumor cells were located in the submucosa but also infiltrated the muscularis propria and the subserosa and were CD34 positive. The molecular study by PCR showed mutation in exon 12 of the PDGFRA gene. IFP is considered a true neoplasm and can also be considered as a potentially invasive lesion.

The prevalence of small-bowel polyps on video capsule endoscopy in patients with sporadic duodenal or ampullary adenomas.

BACKGROUND AND AIMS: Although sporadic duodenal and/or ampullary adenomas (DAs) are uncommon, they are increasingly diagnosed during upper endoscopy. These patients have a 3- to 7-fold increased risk of colonic neoplasia compared with the normal population. It is unknown, however, whether they also have an increased risk of additional small-bowel (SB) polyps. Our aim was to establish the prevalence of SB polyps in patients with DA. METHODS: In a single-center, prospective study, we used video capsule endoscopy (VCE) to investigate the prevalence of SB polyps in patients with a DA compared with patients undergoing VCE for obscure GI bleeding or iron deficiency anemia. RESULTS: Over 25 months, 201 patients were enrolled in the study; the mean age was 65 years and 47% were male. There were 101 control patients and 100 cases of DA cases (mean size, 30 mm (range, 10-80 mm)). We did not identify any SB polyps in either group. Colonic polyps were found more frequently in the DA group compared with controls (61% versus 37%, respectively (P =.002)). Advanced colonic adenoma (high-grade dysplasia, >10 mm, villous histology) were found in 18% of the DA group and 5% of the control group (P =.018). CONCLUSION: Our data suggest that patients with a DA are not at risk for additional SB polyps and hence do not support screening with VCE. However, colonoscopy is mandatory due to the significantly higher risk of colonic polyps including advanced adenomas. (Clinical trial registration number: NCT02470416.).

Application of Deep Learning for Early Screening of Colorectal Precancerous Lesions under White Light Endoscopy.

METHODS: We collected and sorted out the white light endoscopic images of some patients undergoing colonoscopy. The convolutional neural network model is used to detect whether the image contains lesions: CRC, colorectal adenoma (CRA), and colorectal polyps. The accuracy, sensitivity, and specificity rates are used as indicators to evaluate the model. Then, the instance segmentation model is used to locate and classify the lesions on the images containing lesions, and mAP (mean average precision), AP50, and AP75 are used to evaluate the performance of an instance segmentation model. RESULTS: In the process of detecting whether the image contains lesions, we compared ResNet50 with the other four models, that is, AlexNet, VGG19, ResNet18, and GoogLeNet. The result is that ResNet50 performs better than several other models. It scored an accuracy of 93.0%, a sensitivity of 94.3%, and a specificity of 90.6%. In the process of localization and classification of the lesion in images containing lesions by Mask R-CNN, its mAP, AP50, and AP75 were 0.676, 0.903, and 0.833, respectively. CONCLUSION: We developed and compared five models for the detection of lesions in white light endoscopic images. ResNet50 showed the optimal performance, and Mask R-CNN model could be used to locate and classify lesions in images containing lesions.