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

In 2000, the small bowel capsule revolutionized the management of patients with small bowel disorders. Currently, the technological development achieved by the new models of double-headed endoscopic capsules, as miniaturized devices to evaluate the small bowel and colon [pan-intestinal capsule endoscopy (PCE)], makes this non-invasive procedure a disruptive concept for the management of patients with digestive disorders. This technology is expected to identify which patients will require conventional invasive endoscopic procedures (colonoscopy or balloon-assisted enteroscopy), based on the lesions detected by the capsule, i.e., those with an indication for biopsies or endoscopic treatment. The use of PCE in patients with inflammatory bowel diseases, namely Crohn's disease, as well as in patients with iron deficiency anaemia and/or overt gastrointestinal (GI) bleeding, after a non-diagnostic upper endoscopy (esophagogastroduodenoscopy), enables an effective, safe and comfortable way to identify patients with relevant lesions, who should undergo subsequent invasive endoscopic procedures. The recent development of magnetically controlled capsule endoscopy to evaluate the upper GI tract, is a further step towards the possibility of an entirely non-invasive assessment of all the segments of the digestive tract, from mouth-to-anus, meeting the expectations of the early developers of capsule endoscopy.

Detachable string magnetically controlled capsule endoscopy for the noninvasive diagnosis of esophageal diseases: A prospective, blind clinical study.

BACKGROUND: Traditional esophagogastroduodenoscopy (EGD), an invasive examination method, can cause discomfort and pain in patients. In contrast, magnetically controlled capsule endoscopy (MCE), a noninvasive method, is being applied for the detection of stomach and small intestinal diseases, but its application in treating esophageal diseases is not widespread. AIM: To evaluate the safety and efficacy of detachable string MCE (ds-MCE) for the diagnosis of esophageal diseases. METHODS: Fifty patients who had been diagnosed with esophageal diseases were prospectively recruited for this clinical study and underwent ds-MCE and conventional EGD. The primary endpoints included the sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy of ds-MCE for patients with esophageal diseases. The secondary endpoints consisted of visualizing the esophageal and dentate lines, as well as the subjects' tolerance of the procedure. RESULTS: Using EGD as the gold standard, the sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy of ds-MCE for esophageal disease detection were 85.71%, 86.21%, 81.82%, 89.29%, and 86%, respectively. ds-MCE was more comfortable and convenient than EGD was, with 80% of patients feeling that ds-MCE examination was very comfortable or comfortable and 50% of patients believing that detachable string v examination was very convenient. CONCLUSION: This study revealed that ds-MCE has the same diagnostic effects as traditional EGD for esophageal diseases and is more comfortable and convenient than EGD, providing a novel noninvasive method for treating esophageal diseases.

Development and validation of a deep learning system for detection of small bowel pathologies in capsule endoscopy: a pilot study in a Singapore institution.

INTRODUCTION: Deep learning models can assess the quality of images and discriminate among abnormalities in small bowel capsule endoscopy (CE), reducing fatigue and the time needed for diagnosis. They serve as a decision support system, partially automating the diagnosis process by providing probability predictions for abnormalities. METHODS: We demonstrated the use of deep learning models in CE image analysis, specifically by piloting a bowel preparation model (BPM) and an abnormality detection model (ADM) to determine frame-level view quality and the presence of abnormal findings, respectively. We used convolutional neural network-based models pretrained on large-scale open-domain data to extract spatial features of CE images that were then used in a dense feed-forward neural network classifier. We then combined the open-source Kvasir-Capsule dataset (n = 43) and locally collected CE data (n = 29). RESULTS: Model performance was compared using averaged five-fold and two-fold cross-validation for BPMs and ADMs, respectively. The best BPM model based on a pre-trained ResNet50 architecture had an area under the receiver operating characteristic and precision-recall curves of 0.969±0.008 and 0.843±0.041, respectively. The best ADM model, also based on ResNet50, had top-1 and top-2 accuracies of 84.03±0.051 and 94.78±0.028, respectively. The models could process approximately 200-250 images per second and showed good discrimination on time-critical abnormalities such as bleeding. CONCLUSION: Our pilot models showed the potential to improve time to diagnosis in CE workflows. To our knowledge, our approach is unique to the Singapore context. The value of our work can be further evaluated in a pragmatic manner that is sensitive to existing clinician workflow and resource constraints.

Prospective study of diagnostic yields of flexible spectral imaging color enhancement installed in colon capsule endoscopy for colorectal polyps and tumors.

BACKGROUND AND AIMS: We prospectively determined the efficacy of flexible spectral imaging color enhancement (FICE) used with second-generation colon capsule endoscopy (CCE) for colorectal polyps and tumors (CRTs). METHODS: This study included optical colonoscopy within 4 months after CCE. Two colonoscopists independently reviewed CCE using white-light images (CCE-WL) and CCE using FICE images (CCE-FICE), respectively. Based on colonoscopic findings as the criterion standard, the diagnostic accuracy for CRTs was compared between CCE-WL and CCE-FICE. RESULTS: Of 89 enrolled patients (65 men and 24 women; 75 with CRTs including 36 with serrated lesions, 63 with adenomas, and 9 with adenocarcinomas), the per-patient detectability of CCE-FICE for the representative CRTs was significantly higher than that of CCE-WL: overall CRTs (CCE-WL, 79%; CCE-FICE, 88%; P = .0001), 6- to 9-mm CRTs (CCE-WL, 63%; CCE-FICE, 94%; P = .0055), and ≥6-mm CRTs (CCE-WL, 78%; CCE-FICE, 93%; P = .0159). The per-lesion sensitivity of CCE-FICE was significantly higher than that of CCE-WL for CRTs: overall (CCE-WL, 61%; CCE-FICE, 79%; P < .0001), <6 mm (CCE-WL, 53%; CCE-FICE, 69%; P < .0001), 6- to 9-mm CRTs (CCE-WL, 65%; CCE-FICE, 93%; P = .0007), slightly elevated CRTs (CCE-WL, 53%; CCE-FICE, 75%; P < .0001), tubular adenomas (CCE-WL, 61%; CCE-FICE, 79%; P < .0001), and serrated polyps (CCE-WL, 57%; CCE-FICE, 74%; P = .0022). Both modes detected all adenocarcinomas. No significant differences were found between CCE-WL and CCE-FICE of the per-lesion sensitivity for ≥10-mm CRTs (CCE-WL, 81%; CCE-FICE, 94%; P = .1138) or protruding CRTs (CCE-WL, 77%; CCE-FICE, 86%; P = .0614). Kappa coefficients for overall CRTs for CCE-WL and CCE-FICE were .66 and .64, respectively, which indicated substantial agreement. CONCLUSIONS: CCE-FICE improved the detection rates for all CRTs except adenocarcinomas, ≥10-mm polyps, and protruding polyps when compared with CCE-WL. (Clinical trial registration number: UMIN 000021125.).

Comparison of clinical utility of deep learning-based systems for small-bowel capsule endoscopy reading.

BACKGROUND AND AIM: Convolutional neural network (CNN) systems that automatically detect abnormalities from small-bowel capsule endoscopy (SBCE) images are still experimental, and no studies have directly compared the clinical usefulness of different systems. We compared endoscopist readings using an existing and a novel CNN system in a real-world SBCE setting. METHODS: Thirty-six complete SBCE videos, including 43 abnormal lesions (18 mucosal breaks, 8 angioectasia, and 17 protruding lesions), were retrospectively prepared. Three reading processes were compared: (A) endoscopist readings without CNN screening, (B) endoscopist readings after an existing CNN screening, and (C) endoscopist readings after a novel CNN screening. RESULTS: The mean number of small-bowel images was 14 747 per patient. Among these images, existing and novel CNN systems automatically captured 24.3% and 9.4% of the images, respectively. In this process, both systems extracted all 43 abnormal lesions. Next, we focused on the clinical usefulness. The detection rates of abnormalities by trainee endoscopists were not significantly different across the three processes: A, 77%; B, 67%; and C, 79%. The mean reading time of the trainees was the shortest during process C (10.1 min per patient), followed by processes B (23.1 min per patient) and A (33.6 min per patient). The mean psychological stress score while reading videos (scale, 1-5) was the lowest in process C (1.8) but was not significantly different between processes B (2.8) and A (3.2). CONCLUSIONS: Our novel CNN system significantly reduced endoscopist reading time and psychological stress while maintaining the detectability of abnormalities. CNN performance directly affects clinical utility and should be carefully assessed.

Active distance control in multi-capsule endoscopy via closed loop electromagnetic force between capsules.

Capsule endoscopy offers a non-invasive and patient-friendly method for imaging the gastrointestinal tract, boasting superior tissue accessibility compared to traditional endoscopy and colonoscopy. While advances have led to capsules capable of drug delivery, tactile sensing, and biopsy, size constraints often limit a single capsule from having multifunctionality. In response, we introduce multi-capsule endoscopy, where individually ingested capsules, each with unique functionalities, work collaboratively. However, synchronized navigation of these capsules is essential for this approach. In this paper, we present an active distance control strategy using a closed-loop system. This entails equipping one capsule with a sphere permanent magnet and the other with a solenoid. We utilized a Simulink model, incorporating (i) the peristalsis motion on the primary capsule, (ii) a PID controller, (iii) force dynamics between capsules through magnetic dipole approximation, and (iv) position tracking of the secondary capsule. For practical implementation, Hall effect sensors determined the inter-capsule distance, and a PID controller adjusted the solenoid's current to maintain the desired capsule spacing. Our proof-of-concept experiments, conducted on phantoms and ex vivo bovine tissues, pulled the leading capsule mimicking a typical human peristalsis speed of 1 cm/min. Results showcased an inter-capsule distance of 1.94 mm ± 0.097 mm for radii of curvature at 500 mm, 250 mm, and 100 mm, aiming for a 2-mm capsule spacing. For ex vivo bovine tissue, the achieved distance was 0.97 ± 0.28 mm against a target inter-capsule distance of 1 mm. Through the successful demonstration of precise inter-capsule control, this study paves the way for the potential of multi-capsule endoscopy in future research.

Rotational Platform for Real-Time Localization for Active Implantable Medical Devices.

This paper presents a sensor based localization system to localize active implantable medical devices i.e., Wireless Capsule Endoscopy (WCE). The importance of localizing the capsule arises once the images from the capsule detect the abnormalities in the Gastrointestinal tract (GI). A successful system can determine the location that associated with the abnormality for further medical investigation or treatment. The system proposed in this paper comprises a rotational platform that consists of magnetic sensors to detect the position of the embedded magnet in the capsule. The rotational platform provides advantageousness in terms of reducing the number of the sensors and increasing the monitoring accuracy during the real time movement.

Wireless capsule endoscopy multiclass classification using three-dimensional deep convolutional neural network model.

BACKGROUND: Wireless capsule endoscopy (WCE) is a patient-friendly and non-invasive technology that scans the whole of the gastrointestinal tract, including difficult-to-access regions like the small bowel. Major drawback of this technology is that the visual inspection of a large number of video frames produced during each examination makes the physician diagnosis process tedious and prone to error. Several computer-aided diagnosis (CAD) systems, such as deep network models, have been developed for the automatic recognition of abnormalities in WCE frames. Nevertheless, most of these studies have only focused on spatial information within individual WCE frames, missing the crucial temporal data within consecutive frames. METHODS: In this article, an automatic multiclass classification system based on a three-dimensional deep convolutional neural network (3D-CNN) is proposed, which utilizes the spatiotemporal information to facilitate the WCE diagnosis process. The 3D-CNN model fed with a series of sequential WCE frames in contrast to the two-dimensional (2D) model, which exploits frames as independent ones. Moreover, the proposed 3D deep model is compared with some pre-trained networks. The proposed models are trained and evaluated with 29 subject WCE videos (14,691 frames before augmentation). The performance advantages of 3D-CNN over 2D-CNN and pre-trained networks are verified in terms of sensitivity, specificity, and accuracy. RESULTS: 3D-CNN outperforms the 2D technique in all evaluation metrics (sensitivity: 98.92 vs. 98.05, specificity: 99.50 vs. 86.94, accuracy: 99.20 vs. 92.60). In conclusion, a novel 3D-CNN model for lesion detection in WCE frames is proposed in this study. CONCLUSION: The results indicate the performance of 3D-CNN over 2D-CNN and some well-known pre-trained classifier networks. The proposed 3D-CNN model uses the rich temporal information in adjacent frames as well as spatial data to develop an accurate and efficient model.

Reclassification of Inflammatory Bowel Disease Type Unclassified by Small Bowel Capsule Endoscopy.

Background and Objectives: Ulcerative colitis (UC) and Crohn's disease (CD) are idiopathic inflammatory bowel diseases (IBDs) without a unique, gold standard diagnostic test. UC and Crohn's colitis are impossible to distinguish in approximately 10% of cases. The term IBD type unclassified (IBD-U) is recommended for cases of chronic colitis showing overlapping endoscopic, radiological, and biopsy histological features between UC and CD, while indetermined colitis is reserved for colectomy specimens. Our aim was to assess the role of small-bowel capsule endoscopy (SBCE) in the diagnostic work-up of IBD-U. Materials and Methods: We retrospectively studied the cases of IBD-U explored by SBCE in a tertiary referral gastroenterology center. Patients were investigated using SBCE after contraindications were excluded. Diagnostic criteria for small bowel CD consisted in more than three ulcerations, irregular ulcers, or stenosis, and the Lewis score was used for the quantification of inflammation. The immediate impact of reclassification and outcome data was recorded over a follow-up period of more than one year. Results: Twenty-eight patients with IBD-U were examined using SBCE. Nine patients had small bowel lesions that met the diagnostic criteria for CD, resulting in a reclassification rate of 32.1%. In five of these cases, the treatment was subsequently changed. In the remaining nineteen examinations, no significant findings were observed. There were no complications associated with SBCE. Median follow-up time was 32.5 months (range 12-60). During follow-up, twelve patients were classified as having UC, and seven remained as having an unclassified type; one case of colectomy, for medically refractory UC, was recorded. Conclusions: SBCE is a useful safe tool in the work-up of IBD-U, allowing reclassification in about one third of cases, with subsequent treatment modifications. SBCE may provide a definite diagnosis, enhance the comprehension of the disease's progression, and optimize the short- and long-term management strategy.

[Usefulness and safety of observing anorectal lesions by colon capsule endoscopy].

We investigated the findings of rectoanal lesions in 190 patients who underwent colon capsule endoscopy (CCE) at our hospital. Internal hemorrhoids were observed in 70 (36.8%) patients and rectal polyps in 19 (10%) patients. When conventional endoscopy (colonoscopy and double balloon endoscopy) was considered the gold standard, the sensitivity and specificity of rectal polyps were 75% and 93.4%, respectively, and those of internal hemorrhoids were 88.9% and 92.7%, respectively. The prevalence of constipation was significantly higher in the false-negative group for internal hemorrhoids, and the colonic transit time was significantly shorter in the false-negative and false-positive groups for rectal polyps. No adverse events occurred in any of the patients. CCE might be a useful and safe examination method for rectoanal lesions.

Associations between patient factors and successful colon capsule endoscopy - A prospective cohort study.

AIM: To establish patient factors associated with a successful colon capsule endoscopy (CCE) test. METHOD: This prospective cohort study used data collected from patients who underwent CCE as part of the ScotCap evaluation prior to April 2020. A CCE was defined as successful if the capsule visualized the whole colon and rectum (complete test) with sufficient bowel cleansing to permit assessment of the colonic mucosa (adequate bowel preparation). Factors from patients in symptomatic and surveillance groups were analysed for associations with a successful test, complete test, adequate bowel preparation and requirement for further procedure using univariate, multivariate logistic, least absolute shrinkage and selection operator regression. RESULTS: Data from 263 symptomatic and 137 surveillance patients were analysed. There was an association between a symptomatic patient's age and a successful test (OR = 0.97, 95% CI: 0.95-0.99), adequate bowel preparation (OR = 0.97, 95% CI: 0.94-1.00) and further procedure requirement (OR = 1.04, 95% CI: 1.02-1.06). An association was found, for symptomatic patients, between a faecal immunochemical test result in the range 10-399 µg/g and a further procedure (OR = 2.32, 95% CI: 1.23-4.48). In patients undergoing surveillance for previous colorectal cancer (OR = 0.42, 95% CI: 0.18-0.97), had previous bowel resection surgery (OR = 0.43, 95% CI: 0.19-0.98) or took a beta blocker medication (OR = 0.32, 95% CI: 0.11-0.88), an association was found with further procedure requirement. CONCLUSIONS: Among symptomatic patients, younger age was associated with obtaining a successful CCE test. Clinicians could consider patient selection based on these results to improve the rate of successful testing in clinical practice.

What Is the Optimal Bowel Preparation for Capsule Colonoscopy and Pan-intestinal Capsule Endoscopy? A Systematic Review and Meta-Analysis.

BACKGROUND: The rate of adequate cleansing (ACR) and complete examinations (CR) are key quality indicators in capsule colonoscopy (CC) and pan-intestinal capsule endoscopy (PCE). AIMS: To evaluate the efficacy of bowel preparation protocols regarding ACR and CR. METHODS: We conducted a systematic review and meta-analysis, search terms regarding colon capsule preparation, publication date from 2006/01, and date of search 2021/12, in six bibliographic databases. Multiple steps of the cleansing protocol were assessed: diet, adjunctive laxatives, purgative solution, use of prokinetic agents, and "booster". The meta-analytical frequency of ACR and CR was estimated, and subgroup analyses performed. Strategies associated with higher ACR and CR were explored using meta-analytical univariable and multivariable regression models. RESULTS: Twenty-six observational studies and five RCTs included (n = 4072 patients). The pooled rate of ACR was 72.5% (95% C.I. 67.8-77.5%; I2 = 92.4%), and the pooled rate of CR was 83.0% (95% C.I. 78.7-87.7%; I2 = 96.5%). The highest ACR were obtained using a low-fibre diet [78.5% (95% C.I. 72.0-85.6%); I2 = 57.0%], adjunctive laxatives [74.7% (95% C.I. 69.8-80.1%); I2 = 85.3%], and split dose < 4L polyethylene glycol (PEG) as purgative [77.5% (95% C.I. 68.4-87.8%); I2 = 47.3%]. The highest CR were observed using routine prokinetics prior to capsule ingestion [84.4% (95% C.I. 79.9-89.2%); I2 = 89.8%], and sodium phosphate (NaP) as "booster" [86.2% (95% C.I. 82.3-90.2%); I2 = 86.8%]. In univariable models, adjunctive laxatives were associated with higher ACR [OR 1.81 (95% C.I. 1.13; 2.90); p = 0.014]. CR was higher with routine prokinetics [OR 1.86 (95% C.I. 1.13; 3.05); p = 0.015] and split-dose PEG purgative [OR 2.03 (95% C.I. 1.01; 4.09), p = 0.048]. CONCLUSIONS: Main quality outcomes (ACR, CR) remain suboptimal for CC and PCE. Despite considerable heterogeneity, our results support low-fibre diet, use of adjunctive sennosides, split dose < 4L PEG, and routine prokinetics, while NaP remains the most consistent option as booster.

Video capsule endoscopy versus computed tomography enterography in assessing suspected small bowel bleeding: a systematic review and diagnostic test accuracy meta-analysis.

Both computed tomography enterography (CTE) and video capsule endoscopy (VCE) are used in identifying small intestinal pathology in patients with suspected small bowel bleeding (SSBB) following normal upper gastrointestinal endoscopy and colonoscopy. Evidence of the comparative accuracy of these two modalities is crucial for clinical and healthcare decision-making. Comprehensive electronic searches were performed for studies on CTE and/or VCE with reference standard(s). Study selection, data extraction and quality assessment were completed by two authors independently. The QUADAS-2 and QUADAS-C tools were used to assess risk of bias, and applicability. Meta-analysis was performed using a bivariate model to obtain summary estimates of sensitivity, specificity, positive and negative likelihood ratios. Twenty-five studies involving 1986 patients with SSBB were included. Four of these were head-to-head comparison of CTE and VCE. Overall, VCE provided significantly higher sensitivity of 0.74 (95% CI: 0.61-0.83) versus 0.47 (95% CI: 0.32-0.62) for CTE, while CTE showed significantly higher specificity of 0.94 (95% CI: 0.64-0.99) versus 0.53 (95% CI: .36-0.69) for VCE. The positive likelihood ratio of CTE was 7.36 (95% CI: 0.97-56.01) versus 1.58 (95% CI: 1.15-2.15) for VCE and the negative likelihood ratio was 0.49 (95% CI: 0.33-0.72) for VCE versus 0.56 (0.40-0.79) for CTE. A secondary analysis of only head-to-head comparative studies gave results that were similar to the main analysis. Certainty of evidence was moderate. Neither VCE nor CTE is a perfect test for identifying etiology of SSBB in small intestine. VCE was more sensitive while CTE was more specific. Clinicians should choose the appropriate modality depending on whether better sensitivity or specificity is required in each clinical scenario.

Public Imaging Datasets of Gastrointestinal Endoscopy for Artificial Intelligence: a Review.

With the advances in endoscopic technologies and artificial intelligence, a large number of endoscopic imaging datasets have been made public to researchers around the world. This study aims to review and introduce these datasets. An extensive literature search was conducted to identify appropriate datasets in PubMed, and other targeted searches were conducted in GitHub, Kaggle, and Simula to identify datasets directly. We provided a brief introduction to each dataset and evaluated the characteristics of the datasets included. Moreover, two national datasets in progress were discussed. A total of 40 datasets of endoscopic images were included, of which 34 were accessible for use. Basic and detailed information on each dataset was reported. Of all the datasets, 16 focus on polyps, and 6 focus on small bowel lesions. Most datasets (n = 16) were constructed by colonoscopy only, followed by normal gastrointestinal endoscopy and capsule endoscopy (n = 9). This review may facilitate the usage of public dataset resources in endoscopic research.

Jejunal Inflammation in Crohn's Disease: Comparison between Diffusion Weighted Magnetic Resonance Imaging and Video Capsule Endoscopy.

BACKGROUND: Jejunal disease is associated with worse prognosis in Crohn's disease. The added value of diffusion weighted imaging for evaluating jejunal inflammation related to Crohn's Disease is scarce. OBJECTIVES: To compare diffusion weighted imaging, video capsule endoscopy, and inflammatory biomarkers in the assessment of Crohn's disease involving the jejunum. METHODS: Crohn's disease patients in clinical remission were prospectively recruited and underwent magnetic resonance (MR)-enterography and video capsule endoscopy. C-reactive protein and fecal-calprotectin levels were obtained. MR-enterography images were evaluated for restricted diffusion, and apparent diffusion coefficient values were measured. The video capsule endoscopy-based Lewis score was calculated. Associations between diffusion weighted imaging, apparent diffusion coefficient, Lewis score, and inflammatory biomarkers were evaluated. RESULTS: The study included 51 patients, and 27/51 (52.9%) with video capsule endoscopies showed jejunal mucosal inflammation. Sensitivity and specificity of restricted diffusion for video capsule endoscopy mucosal inflammation were 59.3% and 37.5% for the first reader, and 66.7% and 37.5% for the second reader, respectively. Diffusion weighted imaging was not statistically associated with jejunal video capsule endoscopy inflammation (P = 0.813). CONCLUSIONS: Diffusion weighted imaging was not an effective test for evaluation of jejunal inflammation as seen by video capsule endoscopy in patients with quiescent Crohn's disease.

Updates in the diagnosis and management of small-bowel tumors.

Small-bowel tumors represent a rare entity comprising 0.6% of all new cancer cases in the US, and only 3% of all gastrointestinal neoplasms. They are a heterogenous group of neoplasms comprising of about forty different histological subtypes with the most common being adenocarcinoma, neuroendocrine tumors, stromal tumors and lymphomas. Their incidence has been reportedly increasing over recent years, partly owing to the advances and developments in the diagnostic modalities. Small-bowel capsule endoscopy, device assisted enteroscopy and dedicated small-bowel cross-sectional imaging are complimentary tools, supplementing each other in the diagnostic process. Therapeutic management of small-bowel tumors largely depends on the histological type and staging at diagnosis. The aim of the present review article is to discuss relevant advances in the diagnosis and management of small-bowel tumors.

A convolutional neural network for bleeding detection in capsule endoscopy using real clinical data.

BACKGROUND: The goal of the present study was to develop a convolutional neural network for the detection of bleedings in capsule endoscopy videos using realistic clinical data from one single-centre. METHODS: Capsule endoscopy videos from all 133 patients (79 male, 54 female; meanage = 53.73 years, SDage = 26.13) who underwent capsule endoscopy at our institution between January 2014 and August 2018 were screened for pathology. All videos were screened for pathology by two independent capsule experts and confirmed findings were checked again by a third capsule expert. From these videos, 125 pathological findings (individual episodes of bleeding spanning a total of 5696 images) and 103 non-pathological findings (sections of normal mucosal tissue without pathologies spanning a total of 7420 images) were used to develop and validate a neural network (Inception V3) using transfer learning. RESULTS: The overall accuracy of the model for the detection of bleedings was 90.6% [95%CI: 89.4%-91.7%], with a sensitivity of 89.4% [95%CI: 87.6%-91.2%] and a specificity of 91.7% [95%CI: 90.1%-93.2%]. CONCLUSION: Our results show that neural networks can detect bleedings in capsule endoscopy videos under realistic, clinical conditions with an accuracy of 90.6%, potentially reducing reading time per capsule and helping to improve diagnostic accuracy.

Capsule endoscopy diagnosis of gastrointestinal melanoma.

Melanoma is a cancer that frequently metastasises to the small bowel, but most cases are asymptomatic and are diagnosed postmortem. Therefore, CT and PET CT cannot detect all lesions and conventional endoscopic study only detects 10-20% of lesions. In this study, we present the case of a 68-year-old patient with a history of cutaneous melanoma and a diagnosis of intestinal melanoma. Thanks to capsule endoscopy, two lesions compatible with cutaneous melanoma metastasis to the small bowel were detected, allowing a much more effective surgical planning. Capsule endoscopy is an innovative technique that improves preoperative diagnosis, as it is able to detect bowel segments that cannot be inspected by conventional endoscopy. It also has a better resolution than conventional CT, improving sensitivity in the detection of lesions.

Where does capsule endoscopy fit in the diagnostic algorithm of small bowel intussusception?

INTRODUCTION: The investigation of small bowel (SB) intussusception is variable, reflecting the lack of existing standards. The aim of this study was to understand the role of small bowel capsule endoscopy (SBCE) to investigate this pathology. METHODOLOGY: This was a retrospective multi-centre study. Patients with intussusception on SBCE and those where SBCE was carried out due to findings of intussusception on radiological investigations were included. Relevant information was collected. RESULTS: Ninety-five patients (median age 39+/-SD19.1 years, IQR 30) were included. Radiological investigations were carried out in 71 patients (74.7%) prior to SBCE with intussusception being present in 60 patients on radiological investigations (84.5%). Thirty patients (42.2%) had intussusception on radiological investigations followed by a normal SBCE. Ten patients (14.1%) had findings of intussusception on radiological investigations, a normal SBCE and repeat radiological investigations that were also normal. Abnormal findings were noted on SBCE that could explain intussusception on imaging in (16 patients) 22.5% of patients. Five patients (5.3%) underwent radiological investigations and SBCE to investigate coeliac disease and intussusception. None had associated malignancy. Four patients (4.2%) underwent SBCE to investigate familial polyposis syndromes and went on to SB enteroscopy and surgery accordingly. Most patients (n = 14; 14.8%) with intussusception on initial SBCE (without prior radiological imaging) had suspected SB bleeding (n = 10, 10.5%). Four patients (4.2%) had additional findings of a mass on CT scan and went on to have surgery. CONCLUSION: SBCE should be used to complement radiology when investigating intussusception. It is a safe non-invasive test that will minimise unnecessary surgery. Additional radiological investigations following a negative SBCE in cases of intussusception noted on initial radiological investigations are unlikely to yield positive findings. Radiological investigations following intussusception noted on SBCE in case of patients presenting with obscure gastrointestinal bleeding, may yield additional findings.

Evaluating clinical diversity and plausibility of synthetic capsule endoscopic images.

Wireless Capsule Endoscopy (WCE) is being increasingly used as an alternative imaging modality for complete and non-invasive screening of the gastrointestinal tract. Although this is advantageous in reducing unnecessary hospital admissions, it also demands that a WCE diagnostic protocol be in place so larger populations can be effectively screened. This calls for training and education protocols attuned specifically to this modality. Like training in other modalities such as traditional endoscopy, CT, MRI, etc., a WCE training protocol would require an atlas comprising of a large corpora of images that show vivid descriptions of pathologies, ideally observed over a period of time. Since such comprehensive atlases are presently lacking in WCE, in this work, we propose a deep learning method for utilizing already available studies across different institutions for the creation of a realistic WCE atlas using StyleGAN. We identify clinically relevant attributes in WCE such that synthetic images can be generated with selected attributes on cue. Beyond this, we also simulate several disease progression scenarios. The generated images are evaluated for realism and plausibility through three subjective online experiments with the participation of eight gastroenterology experts from three geographical locations and a variety of years of experience. The results from the experiments indicate that the images are highly realistic and the disease scenarios plausible. The images comprising the atlas are available publicly for use in training applications as well as supplementing real datasets for deep learning.