Automatic detection of colorectal polyps using artificial intelligence techniques

Colorectal cancer (CRC) is a leading cause of cancer deaths worldwide. Although polypectomy at early stage reduces CRC incidence, 90% of the polyps are small and diminutive, where removal of them poses risks to patients that may outweigh the benefits. Correctly detecting and predicting polyp type during colonoscopy allows endoscopists to resect and discard the tissue without submitting it for histology, saving time, and costs. Nevertheless, human visual observation of early stage polyps varies. Therefore, this paper aims at developing a fully automatic algorithm to detect and classify hyperplastic and adenomatous colorectal polyps. Adenomatous polyps should be removed, whereas distal diminutive hyperplastic polyps are considered clinically insignificant and may be left in situ . A novel transfer learning application is proposed utilizing features learned from big nonmedical datasets with 1.4-2.5 million images using deep convolutional neural network. The endoscopic images we collected for experiment were taken under random lighting conditions, zooming and optical magnification, including 1104 endoscopic nonpolyp images taken under both white-light and narrowband imaging (NBI) endoscopy and 826 NBI endoscopic polyp images, of which 263 images were hyperplasia and 563 were adenoma as confirmed by histology. The proposed method identified polyp images from nonpolyp images in the beginning followed by predicting the polyp histology. When compared with visual inspection by endoscopists, the results of this study show that the proposed method has similar precision (87.3% versus 86.4%) but a higher recall rate (87.6% versus 77.0%) and a higher accuracy (85.9% versus 74.3%). In conclusion, automatic algorithms can assist endoscopists in identifying polyps that are adenomatous but have been incorrectly judged as hyperplasia and, therefore, enable timely resection of these polyps at an early stage before they develop into invasive cancer.

<div><p class="4"><strong>Objective:</strong> to establish an automatic colonoscopy method based on artificial intelligence. <strong>Methods:</strong> a public database established by a university hospital was used, including colorectal fat and data collection. Initially, all frames in the video are normalized to reduce the high variability between databases. Then, the convolution neural network is used for full depth learning to complete the detection task of polyps. The network starts with the weights learned from millions of natural images in the ImageNet database. According to the fine-tuning technology, the colonoscopy image is used to update the network weight. Finally, the detection of polyps is performed by assigning the probability of containing Po ́ lipo to each table and determining the threshold defined when polyps appears in the table. <strong>Results:</strong> 1875 cases were collected from 5 public databases and databases established by university hospitals, with a total of 123046 forms. The method was trained and evaluated. Comparing the results with the scores of different colonoscopy experts, the accuracy was 0.77, the sensitivity was 0.89, the specificity was 0.71, and the ROC curve (re ceiver operation characteristics) was 0.87. <strong>Conclusion:</strong> compared with experienced gastrointestinal markers, this method overcomes the high variability of different types of lesions and different colonic light conditions (handle, folding or contraction), has very high sensitivity, and can reduce human errors, which is one of the main factors leading to the non detection or leakage of Po lipids in colonoscopy.</p></div>

Over the last few years, researchers have focused on computer‐aided polyp detection in gastroscopy. Deep learning (DL) has shown great promise for polyps' identification. The most exceptional contribution of DL methods in gastroenterology is their ability to identify polyps quickly and accurately using convolution neural network. Nonetheless, despite significant advancements, automatic detection of small polyps remains a challenging and complex task. Furthermore, due to multiple pooling operations, the features of small polyps are lost, resulting in low detection accuracy. This paper proposes an efficient object detection method for polyp detection using gastric images to address this issue. A single‐shot multi‐box detector (SSD) was combined with the feature extractor VGG‐16, and the Refined Map Block (RMB) was integrated into SSD's high‐resolution feature maps to get more semantic information. The RMB output was used as the input to the successive layers. The RMB comprises of attention cascade and feature map concatenation cascade. The attention cascade improved the localization accuracy, while the feature map concatenation cascade improved the classification accuracy. Using the former, the proposed attention‐based SSD for gastric polyps (ASSD‐GPNet) model focused on the specific information, a polyp, rather than the background. Furthermore, the feature map concatenation cascade adds semantic information while reducing computational complexity. The output of these two cascades was combined to produce a refined feature map that enhances the detection of small polyps. The model was trained and tested on 1970 gastric images and Pascal VOC07 + 12. Image augmentation was applied to increase the training data of gastric images to reduce overfitting and skip connections were used to overcome the vanishing gradient problem. Overall, the experimental results demonstrated that the proposed model outperformed than compared models in both medical and natural images. The ASSD‐GPNet obtained mean average precision (mAP) of 94.2% on gastric images and 76.9% on Pascal VOC.

Colorectal cancer (CRC) is one of the most common and deadliest forms of cancer, accounting for nearly 10% of all forms of cancer in the world. Even though colonoscopy is considered the most effective method for screening and diagnosis, the success of the procedure is highly dependent on the operator skills and level of hand-eye coordination. In this work, we propose to adapt fully convolution neural networks (FCN), to identify and segment polyps in colonoscopy images. We converted three established networks into a fully convolution architecture and fine-tuned their learned representations to the polyp segmentation task. We validate our framework on the 2015 MICCAI polyp detection challenge dataset, surpassing the state-of-the-art in automated polyp detection. Our method obtained high segmentation accuracy and a detection precision and recall of 73.61% and 86.31%, respectively.

Colon cancer is thought about as the third most regularly identified cancer after Brest and lung cancer. Most colon cancers are adenocarcinomas developing from adenomatous polyps, grow on the intima of the colon. The standard procedure for polyp detection is colonoscopy, where the success of the standard colonoscopy depends on the colonoscopist experience and other environmental factors. Nonetheless, throughout colonoscopy procedures, a considerable number (8-37%) of polyps are missed due to human mistakes, and these missed polyps are the prospective reason for colorectal cancer cells. In the last few years, many research groups developed deep learning-based computer-aided (CAD) systems that recommended many techniques for automated polyp detection, localization, and segmentation. Still, accurate polyp detection, segmentation is required to minimize polyp miss out rates. This paper suggested a Super-Resolution Generative Adversarial Network (SRGAN) assisted Encoder-Decoder network for fully automated colon polyp segmentation from colonoscopic images. The proposed deep learning model incorporates the SRGAN in the up-sampling process to achieve more accurate polyp segmentation. We examined our model on the publicly available benchmark datasets CVC-ColonDB and Warwick- QU. The model accomplished a dice score of 0.948 on the CVC-ColonDB dataset, surpassed the recently advanced state-of-the-art (SOTA) techniques. When it is evaluated on the Warwick-QU dataset, it attains a Dice Score of 0.936 on part A and 0.895 on Part B. Our model showed more accurate results for sessile and smaller-sized polyps.

Introduction: Bowel preparation is a quality indicator for colonoscopy. Traditionally, Golytely/polyethylene glycol (PEG) has been the main bowel preparation but is not well tolerated by patients due to large volume and taste. Recently, Miralax/Gatorade (M-G) has been introduced as an option to obviate this problem. A meta-analysis concluded that the M-G preparation was inferior to PEG in quality while demonstrating no significant improvement in polyp detection. Our study was performed to examine adenoma detection rate (ADR), rate of one-year repeat colonoscopy due to poor bowel preparation, and rate of non-adherence to the USMSTF (United States Multi-Society Task Force) recommendation for surveillance resulting from suboptimal bowel preparation, given suggestions that low quality bowel preparations can lead to surveillance recommendations inconsistent with guidelines. Methods: Retrospective study examining the above mentioned parameters in 4 consecutive months in 2012, when there was a department-wide change in bowel preparation to M-G, compared with same 4 calendar months in 2011, when PEG was used. Split dosing was used for all patients. Results: 4094 colonoscopies were performed during the study period. After inclusion criteria (average risk adults between ages 50 and 65 years who had screening colonoscopies) and exclusion criteria were applied, 1690 cases remained for analysis. 785 colonoscopies were with M-G and 905 with PEG. There were no statistical differences in age, gender, narcotic usage, and diabetes between two groups. ADR for colonoscopies performed with M-G preparation was 30.5% versus 25.2% for PEG (p=0.02). Proportion of colonoscopies repeated sooner than recommended by USMSTF was 6.4% for M-G and 9.0% for PEG (p=0.05). Proportion of repeat colonoscopies within 1 year due to poor bowel preparation was 3.2% for M-G and 2.4% for PEG (p=0.35). Conclusion: Our retrospective analysis showed higher ADR for M-G compared with PEG. Low quality bowel preparations can result in surveillance recommendations inconsistent with society guidelines, thus adherence to guidelines may be considered a surrogate parameter for bowel preparation quality. In this study, there was a tendency for colonoscopies with PEG to have higher non-adherence to USMSTF recommendations but M-G with more repeat colonoscopies within a year due to poor bowel preparation, though both without statistical significance. More studies are needed to determine the optimal bowel preparation for colonoscopies.

Introduction Adenoma detection rate is important quality indicator in screening colonoscopy. The aim of this study was to measure the change in the colorectal polyp and adenoma detection rate before and after monitoring and feedback in first-year gastroenterology fellow. Method We prospectively analysed a total of 3312 screening colonoscopy data including successful cecal intubation rate, cecal intubation time, withdrawal time, polyp and adenoma detection at the Seoul National University Hospital between March 2013 and September 2014. 1959 colonoscopies performed by 7 first-year gastroenterology fellows without feedback (group A). And 1353 screening colonoscopies performed by another 6 first-year fellows who were informed the periodic monitoring and feedback (group B). We sent the results of polyp and adenoma detection rate to all fellows and professor in charge every month by E-mail. Then we compared before and after feedback group outcomes. Results The average age of patients was 59.4 years, and 55.8% were male. The successful cecal intubation rate was improved after 150 cases both group A (90.0%) and group B (91.3%). Before feedback, overall polyp detection rate was 0.51 ± 0.08, adenoma detection rate was 0.36 ± 0.05. After feedback, the overall polyp detection rate was 0.50 ± 0.05, adenoma detection rate was 0.36 ± 0.03. Compared with before and after feedback group, there was not a significant difference in overall polyp and adenoma detection rate (p = 0.901, p = 0.846). After 150 cases, polyp detection rate was 0.50 for group A, whereas 0.56 for group B (p = 0.041). And adenoma detection rate was significantly higher in group A than group B (0.37 vs. 0.42; p = 0.038). Conclusion The minimum threshold number of technical competence in colonoscopy is more than 150 colonoscopies for trainee who initially started learning colonoscopy. The monitoring and feedback program is associated with an increase in polyp and adenoma detection rate after 150 cases. The periodic monitoring and feedback can be effective for trainees who achieved technical competence. Disclosure of interest None Declared.

Background/Aims:Convolutional neural networks are a class of deep neural networks used for different clinical purposes, including improving the detection rate of colorectal lesions. This systematic review and meta-analysis aimed to assess the performance of convolutional neural network–based models in the detection or classification of colorectal polyps and colorectal cancer.Materials and Methods:A systematic search was performed in MEDLINE, SCOPUS, Web of Science, and other related databases. The performance measures of the convolutional neural network models in the detection of colorectal polyps and colorectal cancer were calculated in the 2 scenarios of the best and worst accuracy. Stata and R software were used for conducting the meta-analysis.Results:From 3368 searched records, 24 primary studies were included. The sensitivity and specificity of convolutional neural network models in predicting colorectal polyps in worst and best scenarios ranged from 84.7% to 91.6% and from 86.0% to 93.8%, respectively. These values in predicting colorectal cancer varied between 93.2% and 94.1% and between 94.6% and 97.7%. The positive and negative likelihood ratios varied between 6.2 and 14.5 and 0.09 and 0.17 in these scenarios, respectively, in predicting colorectal polyps, and 17.1-41.2 and 0.07-0.06 in predicting colorectal polyps. The diagnostic odds ratio and accuracy measures of convolutional neural network models in predicting colorectal polyps in worst and best scenarios ranged between 36% and 162% and between 80.5% and 88.6%, respectively. These values in predicting colorectal cancer in the worst and the best scenarios varied between 239.63% and 677.47% and between 88.2% and 96.4%. The area under the receiver operating characteristic varied between 0.92 and 0.97 in the worst and the best scenarios in colorectal polyps, respectively, and between 0.98 and 0.99 in colorectal polyps prediction.Conclusion:Convolutional neural network–based models showed an acceptable accuracy in detecting colorectal polyps and colorectal cancer.

Application of deep learning for autonomous detection and localization of colorectal polyps in wireless colon capsule endoscopy

Early diagnosis is essential for the successful treatment of bowel cancers including colorectal cancer (CRC), and capsule endoscopic imaging with robotic actuation can be a valuable diagnostic tool when combined with automated image analysis. We present a deep learning rooted detection and segmentation framework for recognizing lesions in colonoscopy and capsule endoscopy images. We restructure established convolution architectures, such as VGG and ResNets, by converting them into fully-connected convolution networks (FCNs), fine-tune them and study their capabilities for polyp segmentation and detection. We additionally use shape-from-shading (SfS) to recover depth and provide a richer representation of the tissue’s structure in colonoscopy images. Depth is incorporated into our network models as an additional input channel to the RGB information and we demonstrate that the resulting network yields improved performance. Our networks are tested on publicly available datasets and the most accurate segmentation model achieved a mean segmentation interception over union (IU) of 47.78% and 56.95% on the ETIS-Larib and CVC-Colon datasets, respectively. For polyp detection, the top performing models we propose surpass the current state-of-the-art with detection recalls superior to 90% for all datasets tested. To our knowledge, we present the first work to use FCNs for polyp segmentation in addition to proposing a novel combination of SfS and RGB that boosts performance.

BackgroundColonic polyps are more likely to be cancerous, especially those with large diameter, large number and atypical hyperplasia. If colonic polyps cannot be treated in early stage, they are likely to develop into colon cancer. Colonoscopy is easily limited by the operator’s experience, and factors such as inexperience and visual fatigue will directly affect the accuracy of diagnosis. Cooperating with Hunan children’s hospital, we proposed and improved a deep learning approach with global average pooling (GAP) in colonoscopy for assisted diagnosis. Our approach for assisted diagnosis in colonoscopy can prompt endoscopists to pay attention to polyps that may be ignored in real time, improve the detection rate, reduce missed diagnosis, and improve the efficiency of medical diagnosis.MethodsWe selected colonoscopy images from the gastrointestinal endoscopy room of Hunan children’s hospital to form the colonic polyp datasets. And we applied the image classification method based on Deep Learning to the classification of Colonic Polyps. The classic networks we used are VGGNets and ResNets. By using global average pooling, we proposed the improved approaches: VGGNets-GAP and ResNets-GAP.ResultsThe accuracies of all models in datasets exceed 98%. The TPR and TNR are above 96 and 98% respectively. In addition, VGGNets-GAP networks not only have high classification accuracies, but also have much fewer parameters than those of VGGNets.ConclusionsThe experimental results show that the proposed approach has good effect on the automatic detection of colonic polyps. The innovations of our method are in two aspects: (1) the detection accuracy of colonic polyps has been improved. (2) our approach reduces the memory consumption and makes the model lightweight. Compared with the original VGG networks, the parameters of our VGG19-GAP networks are greatly reduced.

P-279 - Metabolomics-based test accurately detects and differentiates adenomatous polyps from human sera

Polyps Detection in Colonoscopies

Impact of a real-time automatic quality control system on colorectal polyp and adenoma detection: a prospective randomized controlled study (with videos)

Prospective randomized trials and observational studies have revealed that early detection, classification, and removal of neoplastic colorectal polyp (CP) significantly improve the prevention of colorectal cancer (CRC). The current effectiveness of the diagnostic performance of colonoscopy remains unsatisfactory with unstable accuracy. The convolutional neural networks (CNN) system based on artificial intelligence (AI) technology has demonstrated its potential to help endoscopists in increasing diagnostic accuracy. Nonetheless, several limitations of the CNN system and controversies exist on whether it provides a better diagnostic performance compared to human endoscopists. Therefore, this study sought to address this issue. Online databases (PubMed, Web of Science, Cochrane Library, and EMBASE) were used to search for studies conducted up to April 2020. Besides, the quality assessment of diagnostic accuracy scale-2 (QUADAS-2) was used to evaluate the quality of the enrolled studies. Moreover, publication bias was determined using the Deeks’ funnel plot. In total, 13 studies were enrolled for this meta-analysis (ranged between 2016 and 2020). Consequently, the CNN system had a satisfactory diagnostic performance in the field of CP detection (sensitivity: 0.848 [95% CI: 0.692–0.932]; specificity: 0.965 [95% CI: 0.946–0.977]; and AUC: 0.98 [95% CI: 0.96–0.99]) and CP classification (sensitivity: 0.943 [95% CI: 0.927–0.955]; specificity: 0.894 [95% CI: 0.631–0.977]; and AUC: 0.95 [95% CI: 0.93–0.97]). In comparison with human endoscopists, the CNN system was comparable to the expert but significantly better than the non-expert in the field of CP classification (CNN vs. expert: RDOR: 1.03, P = 0.9654; non-expert vs. expert: RDOR: 0.29, P = 0.0559; non-expert vs. CNN: 0.18, P = 0.0342). Therefore, the CNN system exhibited a satisfactory diagnostic performance for CP and could be used as a potential clinical diagnostic tool during colonoscopy.