Impact of observer variability on the usefulness of endoscopic images for the documentation of upper gastrointestinal endoscopy

LightCF-Net: A Lightweight Long-Range Context Fusion Network for Real-Time Polyp Segmentation

Automatically segmenting polyps from colonoscopy videos is crucial for developing computer-assisted diagnostic systems for colorectal cancer. Existing automatic polyp segmentation methods often struggle to fulfill the real-time demands of clinical applications due to their substantial parameter count and computational load, especially those based on Transformer architectures. To tackle these challenges, a novel lightweight long-range context fusion network, named LightCF-Net, is proposed in this paper. This network attempts to model long-range spatial dependencies while maintaining real-time performance, to better distinguish polyps from background noise and thus improve segmentation accuracy. A novel Fusion Attention Encoder (FAEncoder) is designed in the proposed network, which integrates Large Kernel Attention (LKA) and channel attention mechanisms to extract deep representational features of polyps and unearth long-range dependencies. Furthermore, a newly designed Visual Attention Mamba module (VAM) is added to the skip connections, modeling long-range context dependencies in the encoder-extracted features and reducing background noise interference through the attention mechanism. Finally, a Pyramid Split Attention module (PSA) is used in the bottleneck layer to extract richer multi-scale contextual features. The proposed method was thoroughly evaluated on four renowned polyp segmentation datasets: Kvasir-SEG, CVC-ClinicDB, BKAI-IGH, and ETIS. Experimental findings demonstrate that the proposed method delivers higher segmentation accuracy in less time, consistently outperforming the most advanced lightweight polyp segmentation networks.

Semantic Segmentation of Gastric Polyps in Endoscopic Images Based on Convolutional Neural Networks and an Integrated Evaluation Approach

Convolutional neural networks (CNNs) have received increased attention in endoscopic images due to their outstanding advantages. Clinically, some gastric polyps are related to gastric cancer, and accurate identification and timely removal are critical. CNN-based semantic segmentation can delineate each polyp region precisely, which is beneficial to endoscopists in the diagnosis and treatment of gastric polyps. At present, just a few studies have used CNN to automatically diagnose gastric polyps, and studies on their semantic segmentation are lacking. Therefore, we contribute pioneering research on gastric polyp segmentation in endoscopic images based on CNN. Seven classical semantic segmentation models, including U-Net, UNet++, DeepLabv3, DeepLabv3+, Pyramid Attention Network (PAN), LinkNet, and Muti-scale Attention Net (MA-Net), with the encoders of ResNet50, MobineNetV2, or EfficientNet-B1, are constructed and compared based on the collected dataset. The integrated evaluation approach to ascertaining the optimal CNN model combining both subjective considerations and objective information is proposed since the selection from several CNN models is difficult in a complex problem with conflicting multiple criteria. UNet++ with the MobineNet v2 encoder obtains the best scores in the proposed integrated evaluation method and is selected to build the automated polyp-segmentation system. This study discovered that the semantic segmentation model has a high clinical value in the diagnosis of gastric polyps, and the integrated evaluation approach can provide an impartial and objective tool for the selection of numerous models. Our study can further advance the development of endoscopic gastrointestinal disease identification techniques, and the proposed evaluation technique has implications for mathematical model-based selection methods for clinical technologies.

Pediatric Endoscopy Quality Improvement Network Pediatric Endoscopy Reporting Elements: A Joint NASPGHAN/ESPGHAN Guideline

INTRODUCTION

High-quality procedure reports are a cornerstone of high-quality pediatric endoscopy as they ensure the clear communication of procedural events and outcomes, guide patient care and facilitate continuous quality improvement. The aim of this document is to outline standardized reporting elements that achieved international consensus as requirements for high-quality pediatric endoscopy procedure reports.

METHODS

With support from the North American and European Societies of Pediatric Gastroenterology, Hepatology and Nutrition (NASPGHAN and ESPGHAN), an international working group of the Pediatric Endoscopy Quality Improvement Network (PEnQuIN) used Delphi methodology to identify key elements that should be found in all pediatric endoscopy reports. Item reduction was attained through iterative rounds of anonymized online voting using a 6-point scale. Responses were analyzed after each round and items were excluded from subsequent rounds if ≤50% of panelists rated them as 5 ("agree moderately") or 6 ("agree strongly"). Reporting elements that ≥70% of panelists rated as "agree moderately" or "agree strongly" were considered to have achieved consensus.

RESULTS

Twenty-six PEnQuIN group members from 25 centers internationally rated 63 potential reporting elements that were generated from a systematic literature review and the Delphi panelists. The response rates were 100% for all three survey rounds. Thirty reporting elements reached consensus as essential for inclusion within a pediatric endoscopy report.

DISCUSSION

It is recommended that the PEnQuIN Reporting Elements for pediatric endoscopy be universally employed across all endoscopists, procedures and facilities as a foundational means of ensuring high-quality endoscopy services, while facilitating quality improvement activities in pediatric endoscopy.

Variability in endoscopic assessment of Nissen fundoplication wrap integrity and hiatus herniation

BACKGROUND

Upper endoscopy (EGD) is frequently performed in patients with esophageal complaints following anti-reflux surgery such as fundoplication. Endoscopic evaluation of fundoplication wrap integrity can be challenging. Our primary aim in this pilot study was to evaluate the accuracy and confidence of assessing Nissen fundoplication integrity and hiatus herniation among gastroenterology (GI) fellows, subspecialists, and foregut surgeons.

METHODS

Five variations of post-Nissen fundoplication anatomy were included in a survey of 20 sets of EGD images that was completed by GI fellows, general GI attendings, esophagologists, and foregut surgeons. Accuracy, diagnostic confidence, and inter-rater agreement across providers were evaluated.

RESULTS

There were 31 respondents in the final cohort. Confidence in pre-survey diagnostics significantly differed by provider type (mean confidence out of 5 was 1.8 for GI fellows, 2.7 for general GI attendings, 3.6 for esophagologists, and 3.6 for foregut surgeons, P = 0.01). The mean overall accuracy was 45.9%, which significantly differed by provider type with the lowest rate among GI fellows (37%) and highest among esophagologists (53%; P = 0.01). The accuracy was highest among esophagologists across all wrap integrity variations. Inter-rater agreement was low across wrap integrity variations (Krippendorf's alpha <0.30), indicating low to no agreement between providers.

CONCLUSION

In this multi-center survey study, GI fellows had the lowest accuracy and confidence in assessing EGD images after Nissen fundoplication, whereas esophagologists had the highest. Diagnostic confidence varied considerably and inter-rater agreement was poor. These findings suggest experience may improve confidence, but highlight the need to improve the evaluation of fundoplication wraps.

Morphology of the major papilla predicts ERCP procedural outcomes and adverse events

BACKGROUND

The morphology of the major papilla affects the difficulty of endoscopic retrograde cholangiopancreatography (ERCP), but no associations with adverse events have previously been established. We aimed to assess whether papillary morphology predicts ERCP adverse events.

METHODS

A retrospective analysis was performed of a prospective registry of patients undergoing ERCP for biliary indications. The primary outcome was post-ERCP pancreatitis (PEP), with secondary outcomes including other adverse events and procedural outcomes such as inadvertent pancreatic duct cannulation, cannulation time, and attempts. Papillae were classified as normal (Type I), small or flat (Type II), bulging (Type IIIa), pendulous (Type IIIb), creased (Type IV), or peri-diverticular (Type D). Outcomes were ascertained prospectively at 30 days from index procedures.

RESULTS

A total of 637 patients with native papillae were included. Compared to Type I papillae, Type II and Type IIIb papillae were associated with PEP, with adjusted odds ratios (AOR) of 7.28 (95% confidence intervals, CI, 1.84-28.74) and 4.25 (95% CI 1.26-14.32), respectively. Type II and IIIb papillae were associated with significantly longer cannulation times by 5.37 (95% CI 2.39-8.35) and 4.01 (95% CI 1.72-6.30) minutes, respectively. Type IIIb papillae were associated with lower deep cannulation success (AOR 0.17, 95% CI 0.06-0.48).

CONCLUSION

Papillary morphology is an important factor influencing both ERCP success and outcomes. Understanding this is key for managing intraprocedural approaches and minimizing adverse events.

PROSPECTIVE REGISTRY REGISTRATION

Clinicaltrials.gov identifier NCT04259580.

Taking More Gastroscopy Images Increases the Detection Rate of Clinically Significant Gastric Lesions: Validation of a Systematic Screening Protocol for the Stomach

Background/Aims: For systematic screening protocol for the stomach (SSS), 22 gastroscopy images are considered sufficient to avoid blind spots during gastroscopy. The aim of this study was to investigate the relationship between the number of gastroscopy images taken during the gastroscopy procedure and the detection rate of clinically significant gastric lesions (CSGLs).Materials and Methods: We retrospectively reviewed the data obtained from a cohort of consecutive subjects at a health promotion center. The primary outcome measure was the detection rate of CSGLs per endoscopist, according to the number of gastroscopy images. We also analyzed whether all the CSGLs were detected via SSS.Results: The mean number of gastroscopy images obtained by eight endoscopists was 27.6±10.5 in 2,912 subjects without CSGLs and without biopsies. Among the 5,970 subjects who underwent gastroscopy by the eight endoscopists, 712 CSGLs were detected in 551 subjects. Fifty-six CSGLs (7.9%) in 55 subjects (10.0%) were not detected during the SSS. Photo-endoscopists who took more images achieved a higher detection rate of CSGLs than those who took fewer images (adjusted OR 2.07, 95% CI 1.41~3.05; <i>P</i><0.0001).Conclusions: The modified SSS, which included 22 SSS images, the fundus, and the saddle area, detected significantly more CSGLs. This modified SSS should be validated with further prospective studies.

Recent advances in diagnostic upper endoscopy

BACKGROUND

Esophageo-gastro-duodenoscopy (EGD) is an important procedure used for detection and diagnosis of esophago-gastric lesions. There exists no consensus on the technique of examination.

AIM

To identify recent advances in diagnostic EGDs to improve diagnostic yield.

METHODS

We queried the PubMed database for relevant articles published between January 2001 and August 2019 as well as hand searched references from recently published endoscopy guidelines. Keywords used included free text and MeSH terms addressing quality indicators and technological innovations in EGDs. Factors affecting diagnostic yield and EGD quality were identified and divided into the follow segments: Pre endoscopy preparation, sedation, examination schema, examination time, routine biopsy, image enhanced endoscopy and future developments.

RESULTS

We identified 120 relevant abstracts of which we utilized 67 of these studies in our review. Adequate pre-endoscopy preparation with simethicone and pronase increases gastric visibility. Proper sedation, especially with propofol, increases patient satisfaction after procedure and may improve detection of superficial gastrointestinal lesions. There is a movement towards mandatory picture documentation during EGD as well as dedicating sufficient time for examination improves diagnostic yield. The use of image enhanced endoscopy and magnifying endoscopy improves detection of squamous cell carcinoma and gastric neoplasm. The magnifying endoscopy simple diagnostic algorithm is useful for diagnosis of early gastric cancer.

CONCLUSION

There is a steady momentum in the past decade towards improving diagnostic yield, quality and reporting in EGDs. Other interesting innovations, such as Raman spectroscopy, endocytoscopy and artificial intelligence may have widespread endoscopic applications in the near future.

Quality indicators in digestive endoscopy: introduction to structure, process, and outcome common indicators

The general goal of the project wherein this paper is framed is the proposal of useful quality and safety procedures and indicators to facilitate quality improvement in digestive endoscopy units. This initial offspring sets forth procedures and indicators common to all digestive endoscopy procedures. First, a diagram of pre- and post-digestive endoscopy steps was developed. A group of health care quality and/or endoscopy experts under the auspices of the Sociedad Española de Patología Digestiva (Spanish Society of Digestive Diseases) carried out a qualitative review of the literature regarding the search for quality indicators in endoscopic procedures. Then, a paired analysis was used for the selection of literature references and their subsequent review. Twenty indicators were identified, including seven for structure, eleven for process (five pre-procedure, three intra-procedure, three post-procedure), and two for outcome. Quality of evidence was analyzed for each indicator using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) classification.

Endoscopic classification of the papilla of Vater. Results of an inter- and intraobserver agreement study

BACKGROUND

Many endoscopists acknowledge that the appearance of the papilla of Vater seems to affect biliary cannulation. To assess the association between the macroscopic appearance of the papilla and biliary cannulation and other related clinical issues, a system is needed to define the appearance of the papilla.

OBJECTIVE

The purpose of this study was to validate an endoscopic classification of the papilla of Vater by assessing the interobserver and intraobserver agreements among endoscopist with varying experience.

METHODS

An endoscopic classification, based on pictures captured from 140 different papillae, containing four types of papillae was proposed. The four types are (a) Type 1: regular papilla, no distinctive features, 'classic appearance'; (b) Type 2: small papilla, often flat, with a diameter ≤ 3 mm (approximately 9 Fr); (c) Type 3: protruding or pendulous papilla, a papilla that is standing out, protruding or bulging into the duodenal lumen or sometimes hanging down, pendulous with the orifice oriented caudally; and (d) Type 4: creased or ridged papilla, where the ductal mucosa seems to extend distally, rather out of the papillary orifice, either on a ridge or in a crease. To assess the level of interobserver agreement, a web-based survey was sent out to 18 endoscopists, containing 50 sets of still images of the papilla, distributed between the four different types. Three months later a follow-up survey, with images from the first survey was sent to the same endoscopists.

RESULTS

Interobserver agreement was substantial (κ = 0.62, 95% confidence interval (CI) 0.58-0.65) and were similar for both experts and non-experts. The intraobserver agreement assessed with the second survey was also substantial (κ = 0.66, 95% CI 0.59-0.72).

CONCLUSION

The proposed endoscopic classification of the papilla of Vater seems to be easy to use, irrespective of the level of experience of the endoscopist. It carries a substantial inter- and intraobserver agreement and now the clinical relevance of the four different papilla types awaits to be determined.

Fecal calprotectin one year after ileocaecal resection for Crohn's disease--a comparison with findings at ileocolonoscopy

BACKGROUND AND AIMS

Ileocaecal resection for Crohn's disease is commonly performed. The severity of endoscopic lesions in the anastomotic area one year postoperatively is considered to reflect the subsequent clinical course. Fecal calprotectin (FC) has been shown to correlate with the findings at ileocolonoscopy in Crohn's disease. The objectives of this study were to assess whether the concentration of FC reflects the endoscopic findings one year after ileocaecal resection and to evaluate the variation of FC in individual patients during 6months prior to the ileocolonoscopy.

METHODS

Thirty patients with Crohn's disease and ileocaecal resection performed within one year were included. Stool samples were delivered monthly until an ileocolonoscopy was performed one year postoperatively.

RESULTS

One year after surgery the median values of FC were not significantly different between the patients in endoscopic remission (n=17) and the patients with an endoscopic recurrence (189 (75-364) vs 227 (120-1066)μg/g; p=0.25). However, most patients with low values were in remission and all patients with high (>600μg/g) calprotectin values had recurrent disease. The variability of the FC concentration was most pronounced in patients with diarrhea.

CONCLUSIONS

We found no statistical difference in the concentrations of calprotectin between patients in endoscopic remission and patients with a recurrent disease one year after ileocaecal resection for Crohn's disease. However, among the minority of patients with low or high values, FC indicated remission and recurrence, respectively. There was significant variation of the fecal calprotectin concentrations over time, which affects the utility of calprotectin in clinical practice.

Endosonography For Right-sided and Acute Upper Intestinal Misery: the EFRAIM study: A prospective, randomized, controlled, blinded study

BACKGROUND

Acute upper abdominal pain is a frequent symptom leading to hospital admission.

OBJECTIVE

To determine whether a primary intra- and extraluminal diagnostic approach enabled by endoscopic ultrasound is as effective as a conventional diagnostic algorithm of transabdominal ultrasound followed by oesophagogastroduodenoscopy.

METHODS

A total of 240 patients who presented with acute right-sided and/or upper abdominal pain were prospectively recruited. Exclusion criteria were chronic pain, malignancy, prior abdominal surgery, bleeding, peritonitis, and elevated liver enzymes or lipase as defined 3-times higher than upper reference value. All patients underwent first transabdominal ultrasound and were then randomized (1 : 1) to either endoscopy followed by endoscopic ultrasound or vice versa. Patients and respective examiners were blinded to prior findings.

RESULTS

A total of 223 patients were included. Endoscopic ultrasound provided a higher diagnostic yield than the combination of transabdominal ultrasound and endoscopy (62.3 vs. 50.7%; p = 0.001). For mucosal/intraluminal lesions, we observed a very good agreement between both endoscopic modalities (kappa 0.89). The agreement for pancreatic and biliary causes was good between both ultrasound modalities (kappa 0.66).

CONCLUSIONS

Due to its high diagnostic yield, endoscopic ultrasound as a primary diagnostic modality appears to be a valuable option in patients with acute upper abdominal pain.

Endoscopy reporting standards

OBJECTIVES

The Canadian Association of Gastroenterology (CAG) recently published consensus recommendations for safety and quality indicators in digestive endoscopy. The present article focuses specifically on the identification of key elements that should be found in all electronic endoscopy reports detailing recommendations adopted by the CAG consensus group.

METHODS

A committee of nine individuals steered the CAG Safety and Quality Indicators in Endoscopy Consensus Group, which had a total membership of 35 voting individuals with knowledge on the subject relating to endoscopic services. A comprehensive literature search was performed with regard to the key elements that should be found in an electronic endoscopy report. A task force reviewed all published, full-text, adult and human studies in French or English.

RESULTS

Components to be entered into the standardized report include identification of procedure, timing, procedural personnel, patient demographics and history, indication(s) for procedure, comorbidities, type of bowel preparation, consent for the procedure, pre-endoscopic administration of medications, type and dose of sedation used, extent and completeness of examination, quality of bowel preparation, relevant findings and pertinent negatives, adverse events and resulting interventions, patient comfort, diagnoses, endoscopic interventions performed, details of pathology specimens, details of follow-up arrangements, appended pathology report(s) and, when available, management recommendations. Summary information should be provided to the patient or family.

CONCLUSION

Continuous quality improvement should be the responsibility of every endoscopist and endoscopy facility to ensure improved patient care. Appropriate documentation of endoscopic procedures is a critical component of such activities.

The natural course of Helicobacter pylori infection on endoscopic findings in a population during 17 years of follow-up: the Sørreisa gastrointestinal disorder study

The natural course of Helicobacter pylori (H. pylori) is poorly understood, as most research in the field has been on patient populations. We studied the natural course of H. pylori and its associations to morphological changes of the gastric mucosa, peptic ulcer, and reflux oesophagitis in a prospective cohort study of subjects with and without dyspepsia. A total of 361 adults (201 men/160 women, mean age 41/42 years) in Sørreisa municipality, Norway who in 1987 were subjected to upper endoscopy and assessed for gastrointestinal symptoms and H. pylori status were followed up in 2004. H. pylori was strongly associated with neutrophilic (odds ratio [OR] 23.79; 95% confidence interval [CI] 11.64:48.61) and mononuclear infiltration (OR 9.43; CI 5.12:17.36), moderately with atrophy of the antrum (OR 1.98; CI 1.17:3.34), but not with atrophy of the gastric body or intestinal metaplasia. Elimination of H. pylori was associated with regression of gastric inflammation and atrophy, whereas intestinal metaplasia progressed. H. pylori was positively associated with peptic ulcer (OR 2.69; CI 1.2:6.02) but not significantly negatively associated with oesophagitis (OR 0.62; CI 0.35:1.09). This is the first prospective study including endoscopic findings of subjects without dyspepsia, to show that the impact of H. pylori on gastric atrophy is only modest, and that eliminating H. pylori does not cause regression of intestinal metaplasia. However, inflammation of the gastric mucosa regresses after H. pylori elimination. H. pylori is only a moderate risk factor for peptic ulcer, and other explanatory factors deserve more attention.

A critical review of endoscopic indices in ulcerative colitis: inter-observer variation of the endoscopic index

Various endoscopic indices (EIs) are available for evaluating the endoscopic features of ulcerative colitis (UC), but those currently used in clinical trials are not uniform and have been reported to vary considerably. The currently available EIs for UC include qualitative indices (Baron, Matts and Blackstone scores), the Rachmilewitz endoscopic index (REI; total sum of scores of four types of severity of mucosal lesions), among others. As each of these indices has its own unique characteristics, the selection of the proper EI has to be taken into account in the design of the clinical trial. Endoscopy is an observer-dependent diagnostic method, and inter-observer and intra-observer variations are often problematic. Among other factors, the reliability of clinical trials might be reduced considerably if inter-observer agreement is too low. To date, however, few analyses have focused on inter-observer variation in EIs for UC. The study reported here was undertaken to analyze inter-observer variations in the Baron score and REI. The κ statistics, an indicator of inter-observer agreement, was fair and not particularly high for either the Baron score (κ = 0.31) or REI (κ = 0.27). Modifications made to the REI (modified into a 4-point scale by the authors of this study), making it similar to the Baron score, resulted in κ statistics of 0.44 (moderate). The REI, designed to evaluate many mucosal abnormalities, seemed to enhance the validity of endoscopic grading of UC. It is desirable that inter-observer variations in various EIs are analyzed properly and reviewed comparatively. We emphasize the necessity of establishing the optimal EI for designing clinical trials of UC.