Accuracy of endoscopic diagnosis of Helicobacter pylori infection according to level of endoscopic experience and the effect of training

Predicting Helicobacter pylori infection from endoscopic features

BACKGROUND

Helicobacter pylori infection, prevalent in more than half of the global population, is associated with various gastrointestinal diseases, including peptic ulcers and gastric cancer. The effectiveness of early diagnosis and treatment in preventing gastric cancer highlights the need for improved diagnostic methods. This study aimed to develop a simple scoring system based on endoscopic findings to predict H. pylori infection.

METHODS

A retrospective analysis was conducted on 1,007 patients who underwent upper gastrointestinal endoscopy at Asan Medical Center from January 2019 to December 2021. Exclusion criteria included prior H. pylori treatment, gastric surgery, or gastric malignancies. Diagnostic techniques included rapid urease and 13C-urea breath tests, H. pylori culture, and assessment of endoscopic features following the Kyoto gastritis classification. A new scoring system based on endoscopic findings including regular arrangement of collecting venules (RAC), nodularity, and diffuse or spotty redness was developed for predicting H. pylori infection, utilizing logistic regression analysis in the development set.

RESULTS

The scoring system demonstrated high predictive accuracy for H. pylori infection in the validation set. Scores of 2 and 3 were associated with 96% and 99% infection risk, respectively. Additionally, there was a higher prevalence of diffuse redness and sticky mucus in cases where the initial H. pylori eradication treatment failed.

CONCLUSION

Our scoring system showed potential for improving diagnostic accuracy in H. pylori infection. H. pylori testing should be considered upon spotty redness, diffuse redness, nodularity, and RAC absence on endoscopic findings as determined by the predictive scoring system.

Development and validation of a convolutional neural network model for diagnosing Helicobacter pylori infections with endoscopic images: a multicenter study

BACKGROUND AND AIMS

Insufficient validation limits the generalizability of deep learning in diagnosing Helicobacter pylori infection with endoscopic images. The aim of this study was to develop a deep learning model for the diagnosis of H pylori infection using endoscopic images and validate the model with internal and external datasets.

METHODS

A convolutional neural network (CNN) model was developed based on a training dataset comprising 13,403 endoscopic images from 952 patients who underwent endoscopy at Seoul National University Hospital Gangnam Center. Internal validation was performed using a separate dataset comprised of images of 411 individuals of Korean descent and 131 of non-Korean descent. External validation was performed with the images of 160 patients in Gangnam Severance Hospital. Gradient-weighted class activation mapping was performed to visually explain the model.

RESULTS

In predicting H pylori ever-infected status, the sensitivity, specificity, and accuracy of internal validation for people of Korean descent were .96 (95% confidence interval [CI], .93-.98), .90 (95% CI, .85-.95), and .94 (95% CI, .91-.96), respectively. In the internal validation for people of non-Korean descent, the sensitivity, specificity, and accuracy in predicting H pylori ever-infected status were .92 (95% CI, .86-.98), .79 (95% CI, .67-.91), and .88 (95% CI, .82-.93), respectively. In the external validation cohort, sensitivity, specificity, and accuracy were .86 (95% CI, .80-.93), .88 (95% CI, .79-.96), and .87 (95% CI, .82-.92), respectively, when performing 2-group categorization. Gradient-weighted class activation mapping showed that the CNN model captured the characteristic findings of each group.

CONCLUSIONS

This CNN model for diagnosing H pylori infection showed good overall performance in internal and external validation datasets, particularly in categorizing patients into the never- versus ever-infected groups.

Assessment of Helicobacter pylori infection by deep learning based on endoscopic videos in real time

BACKGROUND AND AIMS

Endoscopic assessment of Helicobacter pylori infection is a simple and effective method. Here, we aimed to develop a deep learning-based system named Intelligent Detection Endoscopic Assistant-Helicobacter pylori (IDEA-HP) to assess H. pylori infection by using endoscopic videos in real time.

METHODS

Endoscopic data were retrospectively obtained from Zhejiang Cancer Hospital (ZJCH) for the development, validation, and testing of the system. Stored videos from ZJCH were used for assessing and comparing the performance of IDEA-HP with that of endoscopists. Prospective consecutive patients undergoing esophagogastroduodenoscopy were enrolled to assess the applicability of clinical practice. The urea breath test was used as the gold standard for diagnosing H. pylori infection.

RESULTS

In 100 videos, IDEA-HP achieved a similar overall accuracy of assessing H. pylori infection to that of experts (84.0% vs. 83.6% [P = 0.729]). Nevertheless, the diagnostic accuracy (84.0% vs. 74.0% [P<0.001]) and sensitivity (82.0% vs. 67.2% [P<0.001]) of IDEA-HP were significantly higher than those of the beginners. In 191 prospective consecutive patients, IDEA-HP achieved accuracy, sensitivity, and specificity of 85.3% (95% CI: 79.0%-89.3%), 83.3% (95% CI: 72.8%-90.5%), and 85.8% (95% CI: 77.7%-91.4%), respectively.

CONCLUSIONS

Our results show that IDEA-HP has great potential for assisting endoscopists in assessing H. pylori infection status during actual clinical work.

An explainable artificial intelligence system for diagnosing Helicobacter Pylori infection under endoscopy: a case-control study

Background

Changes in gastric mucosa caused by Helicobacter pylori (H. pylori) infection affect the observation of early gastric cancer under endoscopy. Although previous researches reported that computer-aided diagnosis (CAD) systems have great potential in the diagnosis of H. pylori infection, their explainability remains a challenge.

Objective

We aim to develop an explainable artificial intelligence system for diagnosing H. pylori infection (EADHI) and giving diagnostic basis under endoscopy.

Design

A case-control study.

Methods

We retrospectively obtained 47,239 images from 1826 patients between 1 June 2020 and 31 July 2021 at Renmin Hospital of Wuhan University for the development of EADHI. EADHI was developed based on feature extraction combining ResNet-50 and long short-term memory networks. Nine endoscopic features were used for H. pylori infection. EADHI's performance was evaluated and compared to that of endoscopists. An external test was conducted in Wenzhou Central Hospital to evaluate its robustness. A gradient-boosting decision tree model was used to examine the contributions of different mucosal features for diagnosing H. pylori infection.

Results

The system extracted mucosal features for diagnosing H. pylori infection with an overall accuracy of 78.3% [95% confidence interval (CI): 76.2-80.3]. The accuracy of EADHI for diagnosing H. pylori infection (91.1%, 95% CI: 85.7-94.6) was significantly higher than that of endoscopists (by 15.5%, 95% CI: 9.7-21.3) in internal test. And it showed a good accuracy of 91.9% (95% CI: 85.6-95.7) in external test. Mucosal edema was the most important diagnostic feature for H. pylori positive, while regular arrangement of collecting venules was the most important H. pylori negative feature.

Conclusion

The EADHI discerns H. pylori gastritis with high accuracy and good explainability, which may improve the trust and acceptability of endoscopists on CADs.

Plain language summary

An explainable AI system for Helicobacter pylori with good diagnostic performance Helicobacter pylori (H. pylori) is the main risk factor for gastric cancer (GC), and changes in gastric mucosa caused by H. pylori infection affect the observation of early GC under endoscopy. Therefore, it is necessary to identify H. pylori infection under endoscopy. Although previous research showed that computer-aided diagnosis (CAD) systems have great potential in H. pylori infection diagnosis, their generalization and explainability are still a challenge. Herein, we constructed an explainable artificial intelligence system for diagnosing H. pylori infection (EADHI) using images by case. In this study, we integrated ResNet-50 and long short-term memory (LSTM) networks into the system. Among them, ResNet50 is used for feature extraction, LSTM is used to classify H. pylori infection status based on these features. Furthermore, we added the information of mucosal features in each case when training the system so that EADHI could identify and output which mucosal features are contained in a case. In our study, EADHI achieved good diagnostic performance with an accuracy of 91.1% [95% confidence interval (CI): 85.7-94.6], which was significantly higher than that of endoscopists (by 15.5%, 95% CI: 9.7-21.3%) in internal test. In addition, it showed a good diagnostic accuracy of 91.9% (95% CI: 85.6-95.7) in external tests. The EADHI discerns H. pylori gastritis with high accuracy and good explainability, which may improve the trust and acceptability of endoscopists on CADs. However, we only used data from a single center to develop EADHI, and it was not effective in identifying past H. pylori infection. Future, multicenter, prospective studies are needed to demonstrate the clinical applicability of CADs.

Artificial intelligence in gastroenterology: A narrative review

Artificial intelligence (AI) is a complex concept, broadly defined in medicine as the development of computer systems to perform tasks that require human intelligence. It has the capacity to revolutionize medicine by increasing efficiency, expediting data and image analysis and identifying patterns, trends and associations in large datasets. Within gastroenterology, recent research efforts have focused on using AI in esophagogastroduodenoscopy, wireless capsule endoscopy (WCE) and colonoscopy to assist in diagnosis, disease monitoring, lesion detection and therapeutic intervention. The main objective of this narrative review is to provide a comprehensive overview of the research being performed within gastroenterology on AI in esophagogastroduodenoscopy, WCE and colonoscopy.

A novel endoscopic finding of a scratch sign is useful for evaluating the Helicobacter pylori infection status

Objectives

During esophagogastroduodenoscopy, a red linear scrape-like appearance with white deposits sometimes appears on the gastric mucosa at the lower greater curvature of the gastric body, a finding we named the "scratch sign." We aimed to clarify the clinical significance of this new endoscopic finding in the endoscopic evaluation of the Helicobacter pylori infection status.

Methods

Among patients who underwent esophagogastroduodenoscopy at our hospital between October 2016 and June 2017, 437 patients were included in the study. We first examined the overall scratch sign positivity rate, and then this was compared according to the H. pylori infection status. Subsequently, other variables were compared and examined between the positive and negative scratch sign groups.

Results

Overall, 437 patients were included in the analysis. The scratch sign was observed in 1.4% of 71 patients with current infections, 26.9% of 290 patients with past infections, and 31.6% of 76 uninfected patients. In the multivariate analysis, H. pylori-negative, severe gastric mucosal atrophy, and acid secretion depressant were independent factors that significantly affected the appearance of the scratch sign.

Conclusions

A novel endoscopic finding, the scratch sign, was found to be a good endoscopic predictor of H. pylori-negative gastric mucosa. Furthermore, combined with atrophic changes and xanthomas that persisted after eradication, these findings were found to be useful in accurately diagnosing H. pylori past-infected gastric mucosa endoscopically.

Current Status of Artificial Intelligence-Based Computer-Assisted Diagnosis Systems for Gastric Cancer in Endoscopy

Artificial intelligence (AI) is gradually being utilized in various fields as its performance has been improving with the development of deep learning methods, availability of big data, and the progression of computer processing units. In the field of medicine, AI is mainly implemented in image recognition, such as in radiographic and pathologic diagnoses. In the realm of gastrointestinal endoscopy, although AI-based computer-assisted detection/diagnosis (CAD) systems have been applied in some areas, such as colorectal polyp detection and diagnosis, so far, their implementation in real-world clinical settings is limited. The accurate detection or diagnosis of gastric cancer (GC) is one of the challenges in which performance varies greatly depending on the endoscopist's skill. The diagnosis of early GC is especially challenging, partly because early GC mimics atrophic gastritis in the background mucosa. Therefore, several CAD systems for GC are being actively developed. The development of a CAD system for GC is considered challenging because it requires a large number of GC images. In particular, early stage GC images are rarely available, partly because it is difficult to diagnose gastric cancer during the early stages. Additionally, the training image data should be of a sufficiently high quality to conduct proper CAD training. Recently, several AI systems for GC that exhibit a robust performance, owing to being trained on a large number of high-quality images, have been reported. This review outlines the current status and prospects of AI use in esophagogastroduodenoscopy (EGDS), focusing on the diagnosis of GC.

Artificial intelligence in gastric cancer: applications and challenges

Gastric cancer (GC) is one of the most common malignant tumors with high mortality. Accurate diagnosis and treatment decisions for GC rely heavily on human experts' careful judgments on medical images. However, the improvement of the accuracy is hindered by imaging conditions, limited experience, objective criteria, and inter-observer discrepancies. Recently, the developments of machine learning, especially deep-learning algorithms, have been facilitating computers to extract more information from data automatically. Researchers are exploring the far-reaching applications of artificial intelligence (AI) in various clinical practices, including GC. Herein, we aim to provide a broad framework to summarize current research on AI in GC. In the screening of GC, AI can identify precancerous diseases and assist in early cancer detection with endoscopic examination and pathological confirmation. In the diagnosis of GC, AI can support tumor-node-metastasis (TNM) staging and subtype classification. For treatment decisions, AI can help with surgical margin determination and prognosis prediction. Meanwhile, current approaches are challenged by data scarcity and poor interpretability. To tackle these problems, more regulated data, unified processing procedures, and advanced algorithms are urgently needed to build more accurate and robust AI models for GC.

Linked Color Imaging (LCI) Emphasizes the Color Changes in the Gastric Mucosa After Helicobacter pylori Eradication

BACKGROUND

Diffuse redness is a characteristic endoscopic finding that indicates current infection of Helicobacter pylori, which is reduced after successful eradication. Linked color imaging (LCI) has been reported to improve the visibility of diffuse redness compared to white light imaging (WLI); however, quantitative evaluation has not been reported.

AIMS

This study aimed to objectively evaluate the color change of the gastric mucosa after H. pylori eradication.

METHODS

Images of the greater curvature of the antrum and corpus were captured, and the sites were biopsied during esophagogastroduodenoscopy (EGD) before and 1 year after eradication. The region of interest (ROI) was set around the biopsied area on the images. The color difference (ΔE) before and after eradication was calculated using the CIE L*a*b* color space. The association between the histological evaluation and the color value of the corresponding ROI was determined.

RESULTS

At the antrum, there was no significant color change with either mode. At the corpus, the a* value, which reflected redness, decreased significantly after eradication with both modes (WLI: 41.2 to 36.0, LCI: 37.5 to 25.5); the b* value, reflecting yellowish, decreased with WLI, but increased significantly with LCI (WLI: 44.6 to 41.6, LCI: 23.9 to 29.2). The ΔE was significantly larger with LCI than with WLI (16.5 vs. 8.6). The a* values at the corpus were generally associated with histological neutrophil infiltration.

CONCLUSIONS

Quantitative evaluation revealed that LCI emphasizes the change in color of the gastric mucosa due to the reduction in diffuse redness.

Artificial intelligence versus expert endoscopists for diagnosis of gastric cancer in patients who have undergone upper gastrointestinal endoscopy

AIMS

To compare endoscopy gastric cancer images diagnosis rate between artificial intelligence (AI) and expert endoscopists.

PATIENTS AND METHODS

We used the retrospective data of 500 patients, including 100 with gastric cancer, matched 1:1 to diagnosis by AI or expert endoscopists. We retrospectively evaluated the noninferiority (prespecified margin 5 %) of the per-patient rate of gastric cancer diagnosis by AI and compared the per-image rate of gastric cancer diagnosis.

RESULTS

Gastric cancer was diagnosed in 49 of 49 patients (100 %) in the AI group and 48 of 51 patients (94.12 %) in the expert endoscopist group (difference 5.88, 95 % confidence interval: -0.58 to 12.3). The per-image rate of gastric cancer diagnosis was higher in the AI group (99.87 %, 747 /748 images) than in the expert endoscopist group (88.17 %, 693 /786 images) (difference 11.7 %).

CONCLUSIONS

Noninferiority of the rate of gastric cancer diagnosis by AI was demonstrated but superiority was not demonstrated.

Artificial Intelligence in Endoscopy

Artificial intelligence (AI) is rapidly developing in various medical fields, and there is an increase in research performed in the field of gastrointestinal (GI) endoscopy. In particular, the advent of convolutional neural network, which is a class of deep learning method, has the potential to revolutionize the field of GI endoscopy, including esophagogastroduodenoscopy (EGD), capsule endoscopy (CE), and colonoscopy. A total of 149 original articles pertaining to AI (27 articles in esophagus, 30 articles in stomach, 29 articles in CE, and 63 articles in colon) were identified in this review. The main focuses of AI in EGD are cancer detection, identifying the depth of cancer invasion, prediction of pathological diagnosis, and prediction of Helicobacter pylori infection. In the field of CE, automated detection of bleeding sites, ulcers, tumors, and various small bowel diseases is being investigated. AI in colonoscopy has advanced with several patient-based prospective studies being conducted on the automated detection and classification of colon polyps. Furthermore, research on inflammatory bowel disease has also been recently reported. Most studies of AI in the field of GI endoscopy are still in the preclinical stages because of the retrospective design using still images. Video-based prospective studies are needed to advance the field. However, AI will continue to develop and be used in daily clinical practice in the near future. In this review, we have highlighted the published literature along with providing current status and insights into the future of AI in GI endoscopy.

Deep learning for gastroscopic images: computer-aided techniques for clinicians

Gastric disease is a major health problem worldwide. Gastroscopy is the main method and the gold standard used to screen and diagnose many gastric diseases. However, several factors, such as the experience and fatigue of endoscopists, limit its performance. With recent advancements in deep learning, an increasing number of studies have used this technology to provide on-site assistance during real-time gastroscopy. This review summarizes the latest publications on deep learning applications in overcoming disease-related and nondisease-related gastroscopy challenges. The former aims to help endoscopists find lesions and characterize them when they appear in the view shed of the gastroscope. The purpose of the latter is to avoid missing lesions due to poor-quality frames, incomplete inspection coverage of gastroscopy, etc., thus improving the quality of gastroscopy. This study aims to provide technical guidance and a comprehensive perspective for physicians to understand deep learning technology in gastroscopy. Some key issues to be handled before the clinical application of deep learning technology and the future direction of disease-related and nondisease-related applications of deep learning to gastroscopy are discussed herein.

Diagnostic Accuracy of H. pylori Status by Conventional Endoscopy: Time-Trend Change After Eradication and Impact of Endoscopic Image Quality

Aim

To assess the time trend of diagnostic accuracy of pre- and post-eradication H. pylori status and interobserver agreement of gastric atrophy grading.

Methods

A series 100 of conventional endoscopic image sets taken from subjects undergoing gastric cancer screening at a polyclinic were evaluated by 5 experienced assessors. Each assessor independently examined endoscopic findings according to the Kyoto classification and then determined the H. pylori status (never, current, or past infected). Gastric atrophy was assessed according to the Kimura-Takemoto classification and classified into 3 grades (none/mild, moderate, or severe). The image series that ≥3 assessors considered to have good quality were arbitrarily defined as high-quality image (HQI) series, and the rest were defined as low-quality image (LQI) series.

Results

The overall diagnostic accuracy of H. pylori status was 83.0%. It was lowest in subjects with current infection (54%), gradually increased at 1 year (79%, P < 0.001) and 3 years (94.0%, P = 0.002), but then did not significantly change at 5 years (91.0%, P = 0.420) after eradication. The rate of LQI series was 28%. The overall diagnostic accuracy of H. pylori status dropped from 88.9% to 67.9% (P < 0.001), and the mean kappa value on gastric atrophy grading dropped from 0.730 to 0.580 (P = 0.002) in the HQI and LQI series, respectively.

Conclusions

Diagnostic accuracy of H. pylori status increased over time after eradication. LQI series badly affected the diagnostic accuracy of H. pylori status and the level of agreement when grading gastric atrophy.

A New Dawn for the Use of Artificial Intelligence in Gastroenterology, Hepatology and Pancreatology

Artificial intelligence (AI) is rapidly becoming an essential tool in the medical field as well as in daily life. Recent developments in deep learning, a subfield of AI, have brought remarkable advances in image recognition, which facilitates improvement in the early detection of cancer by endoscopy, ultrasonography, and computed tomography. In addition, AI-assisted big data analysis represents a great step forward for precision medicine. This review provides an overview of AI technology, particularly for gastroenterology, hepatology, and pancreatology, to help clinicians utilize AI in the near future.

Diagnosis of Helicobacter pylori Infection by the Arrangement of Collecting Venules Using White Light Endoscopy: Evaluation of Interobserver Agreement

BACKGROUND

Regular arrangement of collecting venules (RAC) in gastric mucosa accurately identifies patients without Helicobacter pylori (H pylori) infection. The aim of our study was to evaluate the reproducibility of RAC using white light endoscopy without magnification, in a European country, and to assess the impact of proton pump inhibitors (PPIs).

METHODS

A multicenter prospective study with image capture of the distal lesser gastric curvature and gastric biopsies was performed. The presence of starfish-like minute points regularly distributed throughout lesser curvature was considered as RAC positive (RAC+). A set of 20 images was used for the training phase and inter and intra-observer agreements were calculated.

RESULTS

174 patients were included and 85 (48.9%) were taking PPIs. Kappa values for interobserver and intra-observer agreements were substantial (0.786) and excellent (0.906), respectively. H. pylori infection was diagnosed in 29 patients (16.7%): 10/85 with PPIs and 19/89 without PPIs (11.8% vs. 21.3%; p = 0.09). All RAC + patients were free of H. pylori infection, with a sensitivity and negative predictive value of 100%, regardless of PPI intake.

CONCLUSION

The endoscopic diagnosis of H. pylori by RAC is an easy-to-learn and highly reproducible technique, even with PPI intake. Our results warrant RAC as a real-time diagnostic method for H. pylori-negative infection in Western practice.

Current situation and prospect of artificial intelligence application in endoscopic diagnosis of Helicobacter pylori infection

With the appearance and prevalence of deep learning, artificial intelligence (AI) has been broadly studied and made great progress in various fields of medicine, including gastroenterology. Helicobacter pylori (H. pylori), closely associated with various digestive and extradigestive diseases, has a high infection rate worldwide. Endoscopic surveillance can evaluate H. pylori infection situations and predict the risk of gastric cancer, but there is no objective diagnostic criteria to eliminate the differences between operators. The computer-aided diagnosis system based on AI technology has demonstrated excellent performance for the diagnosis of H. pylori infection, which is superior to novice endoscopists and similar to skilled. Compared with the visual diagnosis of H. pylori infection by endoscopists, AI possesses voluminous advantages: High accuracy, high efficiency, high quality control, high objectivity, and high-effect teaching. This review summarizes the previous and recent studies on AI-assisted diagnosis of H. pylori infection, points out the limitations, and puts forward prospect for future research.

Current status and future perspective of artificial intelligence applications in endoscopic diagnosis and management of gastric cancer

Image recognition using artificial intelligence (AI) has progressed significantly due to innovative technologies such as machine learning and deep learning. In the field of gastric cancer (GC) management, research on AI-based diagnosis such as anatomical classification of endoscopic images, diagnosis of Helicobacter pylori infection, and detection and qualitative diagnosis of GC is being conducted, and an accuracy equivalent to that of physicians has been reported. It is expected that AI will soon be introduced in the field of endoscopic diagnosis and management of gastric cancer as a supportive tool for physicians, thus improving the quality of medical care.

Accuracy of Endoscopic Diagnosis of Helicobacter pylori Based on the Kyoto Classification of Gastritis: A Multicenter Study

Background

There is lack of clinical evidence supporting the value of the Kyoto classification of gastritis for the diagnosis of Helicobacter pylori (H. pylori) infection in Chinese patients, and there aren’t enough specific features for the endoscopic diagnosis of past infections, which is of special significance for the prevention of early gastric cancer (GC).

Methods

This was a prospective and multicenter study with 650 Chinese patients. The H. pylori status and gastric mucosal features, including 17 characteristics based on the Kyoto classification and two newly-defined features unclear atrophy boundary (UAB) and RAC reappearance in atrophic mucosa (RAC reappearance) were recorded in a blind fashion. The clinical characteristics of the subjects were analyzed, and the diagnostic odds ratio (DOR), sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), area under the receiver operating characteristics curve (ROC/AUC), and 95% confidence intervals were calculated for the different features, individually, and in combination.

Results

For past infection, the DOR of UAB was 7.69 (95%CI:3.11−19.1), second only to map-like redness (7.78 (95%CI: 3.43−17.7)). RAC reappearance showed the highest ROC/AUC (0.583). In cases in which at least one of these three specific features of past infection was considered positive, the ROC/AUC reached 0.643. For current infection, nodularity showed the highest DOR (11.7 (95%CI: 2.65−51.2)), followed by diffuse redness (10.5 (95%CI: 4.87−22.6)). Mucosal swelling showed the highest ROC/AUC (0.726). Regular arrangement of collecting venules (RAC) was specific for no infection.

Conclusions

This study provides evidence of the clinical accuracy and robustness of the Kyoto classification of gastritis for the diagnosis of H. pylori in Chinese patients, and confirms UAB and RAC reappearance partly supplement it for the diagnosis of past infections, which is of great benefit to the early prevention of GC.

Endoscopic three-categorical diagnosis of Helicobacter pylori infection using linked color imaging and deep learning: a single-center prospective study (with video)

BACKGROUND

Helicobacter pylori (H. pylori) eradication is required to reduce incidence related to gastric cancer. Recently, it was found that even after the successful eradication of H. pylori, an increased, i.e., moderate, risk of gastric cancer persists in patients with advanced mucosal atrophy and/or intestinal metaplasia. This study aimed to develop a computer-aided diagnosis (CAD) system to classify the status of H. pylori infection of patients into three categories: uninfected (with no history of H. pylori infection), currently infected, and post-eradication.

METHODS

The CAD system was based on linked color imaging (LCI) combined with deep learning (DL). First, a validation dataset was formed for the CAD systems by recording endoscopic movies of 120 subjects. Next, a training dataset of 395 subjects was prepared to enable DL. All endoscopic examinations were recorded using both LCI and white-light imaging (WLI). These endoscopic data were used to develop two different CAD systems, one for LCI (LCI-CAD) and one for WLI (WLI-CAD) images.

RESULTS

The diagnostic accuracy of the LCI-CAD system was 84.2% for uninfected, 82.5% for currently infected, and 79.2% for post-eradication status. Comparisons revealed superior accuracy of diagnoses based on LCI-CAD data relative based on WLI-CAD for uninfected, currently infected, and post-eradication cases. Furthermore, the LCI-CAD system demonstrated comparable diagnostic accuracy to that of experienced endoscopists with the validation data set of LCI.

CONCLUSIONS

The results of this study suggest the feasibility of an innovative gastric cancer screening program to determine cancer risk in individual subjects based on LCI-CAD.

The endoscopic predictors of Helicobacter pylori status: a meta-analysis of diagnostic performance

Objective:

The endoscopic findings associated with Helicobacter pylori–naïve status, current infection or past infection are an area of ongoing interest. Previous studies have investigated parameters with a potential diagnostic value. The aim of this study was to perform meta-analysis of the available literature to validate the diagnostic accuracy of mucosal features proposed in the Kyoto classification.

Data sources:

The databases of MEDLINE and Embase, clinicalTrials.gov and the Cochrane Library were systematically searched for relevant studies from October 1999 to October 2019.

Methods:

A bivariate random effects model was used to produce pooled diagnostic accuracy calculations for each of the studied endoscopic findings. Diagnostic odds ratios and sensitivity and specificity characteristics were calculated to identify significant predictors of H pylori status.

Results:

Meta-analysis included 4380 patients in 15 studies. The most significant predictor of an H pylori-naïve status was a regular arrangement of collecting venules (diagnostic odds ratio 55.0, sensitivity 78.3%, specificity 93.8%). Predictors of active H pylori infection were mucosal oedema (18.1, 63.7%, 91.1%) and diffuse redness (14.4, 66.5%, 89.0%). Map-like redness had high specificity for previous H pylori eradication (99.0%), but poor specificity (13.0%).

Conclusion:

The regular arrangement of collecting venules, mucosal oedema, diffuse redness and map-like redness are important endoscopic findings for determining H pylori status. This meta-analysis provides a tentative basis for developing future endoscopic classification systems.

Optimized diagnosis of Helicobacter pylori and tailored eradication therapy for preventing gastric cancer: a proposal for SHAKE strategy

INTRODUCTION

To decrease gastric cancer-related mortality, the Korean National Cancer Screening Program provides biennial screening gastroscopy to all individuals aged >40 years. However, a test-and-treat strategy of Helicobacter pylori for preventing gastric cancer has not been established.

AREAS COVERED

In this review, we present up-to-date results of endoscopic findings of H. pylori gastritis, optimal sites for H. pylori detection, gastric cancer risk assessment using serum pepsinogen, tailored eradication based on the antimicrobial resistance against H. pylori, and post-eradication surveillance.

EXPERT OPINION

Here we propose approaches to H. pylori diagnosis and treatment for preventing gastric cancer, termed 'Screening for H. pylori in Korea and Eradication (SHAKE)' strategy. This strategy consists of the following: (1) optimized H. pylori diagnosis, (2) individualized management based on the H. pylori infection status, and (3) tailored eradication therapy. H. pylori gastritis can be diagnosed by endoscopic observation of the gastric mucosal pattern at the greater curvature of the corpus. Measurement of the serum pepsinogen I/II ratio is useful for assessing the risk of gastric cancer. As a first-line treatment, tailored eradication based on the results of molecular testing is effective in a country with a high rate of clarithromycin-resistant H. pylori.

A systematic review of the role of non-magnified endoscopy for the assessment of H. pylori infection

Background and study aims  There is growing interest in the endoscopic recognition of Helicobacter pylori infection, and application to routine practice. We present a systematic review of the current literature regarding diagnosis of H. pylori during standard (non-magnified) endoscopy, including adjuncts such as image enhancement and computer-aided diagnosis. Method  The Medline and Cochrane databases were searched for studies investigating performance of non-magnified optical diagnosis for H. pylori , or those which characterized mucosal features associated with H. pylori infection. Studies were preferred with a validated reference test as the comparator, although they were included if at least one validated reference test was used. Results  Twenty suitable studies were identified and included for analysis. In total, 4,703 patients underwent investigation including white light endoscopy, narrow band imaging, i-scan, blue-laser imaging, and computer-aided diagnostic techniques. The endoscopic features of H. pylori infection observed using each modality are discussed and diagnostic accuracies reported. The regular arrangement of collecting venules (RAC) is an important predictor of the H. pylori -naïve stomach. "Mosaic" and "mottled" patterns have a positive association with H. pylori infection. The "cracked" pattern may be a predictor of an H. pylori- negative stomach following eradication. Conclusions  This review summarizes current progress made in endoscopic diagnosis of H. pylori infection. At present there is no single diagnostic approach that provides validated diagnostic accuracy. Further prospective studies are required, as is development of a validated classification system. Early studies in computer-aided diagnosis suggest potential for a high level of accuracy but real-time results are awaited.

Validity of endoscopic features for the diagnosis of Helicobacter pylori infection status based on the Kyoto classification of gastritis

OBJECTIVES

Evaluation of Helicobacter pylori infection status (non-infection, past infection, current infection) has become important. This study aimed to determine the usefulness of the Kyoto classification of gastritis for diagnosing H. pylori infection status by endoscopy.

METHODS

In this prospective study, 498 subjects were recruited. Seven well-experienced endoscopists blinded to the history of eradication therapy performed the examinations. Endoscopic findings were assessed according to the Kyoto classification of gastritis: diffuse redness, regular arrangement of collecting venules (RAC), fundic gland polyp (FGP), atrophy, xanthoma, hyperplastic polyp, map-like redness, intestinal metaplasia, nodularity, mucosal swelling, white and flat elevated lesion, sticky mucus, depressive erosion, raised erosion, red streak, and enlarged folds. We established prediction models according to a machine learning procedure and compared them with general assessment by endoscopists using the Kyoto classification of gastritis.

RESULTS

Significantly higher diagnostic odds were obtained for RAC (32.2), FGP (7.7), and red streak (4.7) in subjects with non-infection, map-like redness (12.9) in subjects with past infection, and diffuse redness (26.8), mucosal swelling (13.3), sticky mucus (10.2) and enlarged fold (8.6) in subjects with current infection. The overall diagnostic accuracy rate was 82.9% with the Kyoto classification of gastritis. The diagnostic accuracy of the prediction model was 88.6% for the model without H. pylori eradication history and 93.4% for the model with eradication history.

CONCLUSIONS

The Kyoto classification of gastritis is useful for diagnosing H. pylori infection status based on endoscopic findings. Our prediction model is helpful for novice endoscopists. (UMIN000016674).

Association between a regular arrangement of collecting venules and absence of Helicobacter pylori infection in a European population

BACKGROUND AND AIMS

Helicobacter pylori is the major cause of gastritis and gastritis-associated diseases. Detection of a regular arrangement of collecting venules (RAC) in the lesser gastric curvature correlates with negative H pylori status with a sensitivity and negative predictive value (NPV) higher than 90% in Asian countries. The aim of the study was to evaluate the value of RAC as a diagnostic method of H pylori infection in a European population.

METHODS

A prospective study with high-definition endoscopes without magnification was performed by 3 endoscopists. The presence of starfish-like minute points regularly distributed throughout the lesser curvature of the gastric body was considered RAC positive (RAC+). Gastric biopsies were performed during the procedure for H pylori diagnosis.

RESULTS

One hundred forty patients were included from February 2017 to May 2018. The prevalence of H pylori infection was 31% and 47 of 140 patients (34%) were RAC+; 13 of 23 patients in whom H pylori was eradicated were RAC+. The mean age of RAC+ patients was lower (44.4 vs 52.4 years, P = .004) and they had less- significant endoscopic findings (9; 19.1% vs 38; 80.9%; P = .017). Gender, use of nonsteroidal anti-inflammatory drugs, antithrombotic or anticoagulants treatments, and a history of H pylori eradication did not show differences in the RAC pattern. The absence of RAC was associated with H pylori infection in 47.3% (44/93) of cases. In contrast, all RAC+ patients were free of H pylori infection, with sensitivity and NPV of 100% for the exclusion of H pylori infection.

CONCLUSION

The presence of RAC+ in the lesser curvature evaluated with high-definition endoscopy can accurately identify patients without H pylori.

Factors associated with oesophagogastric cancers missed by gastroscopy: a case-control study

INTRODUCTION

There is increasing demand for gastroscopy in the United Kingdom. In around 10% of patients, gastroscopy is presumed to have missed oesophagogastric (OG) cancer prior to diagnosis. We examine patient, endoscopist and service level factors that may affect rates of missed OG cancers.

METHODS

Gastroscopies presumed to have missed OG cancers performed up to 3 years prior to diagnosis were identified over 6 years in Sheffield, UK. Factors related to the patient, endoscopist and endoscopy lists were examined in a case-control study. Procedures which missed cancer were compared with two procedure controls: the procedures which subsequently diagnosed cancer in the same patient, and second, endoscopist matched procedures diagnostic of small benign focal lesions.

RESULTS

We identified 48 (7.7%) cases of missed OG cancer. Endoscopy lists on which OG cancer diagnoses were missed contained a greater number of total procedures compared with lists on which diagnoses were subsequently made (OR 1.42 95% CI 1.13 to 1.78) and when compared with lists during which matched endoscopists diagnosed benign small focal lesions (OR 1.25, 95% CI 1.02 to 1.52). The use of sedation, endoscopist profession and experience, or time of procedure were not associated with a missed cancer.

CONCLUSION

7.7% of patients diagnosed with OG cancer could have been diagnosed and treated earlier. Our study suggests that endoscopy lists with greater numbers of procedures may be associated with missed OG cancers. The use of sedation, endoscopist background or time of procedure did not increase the risk of missed cancer procedures.

Artificial intelligence and upper gastrointestinal endoscopy: Current status and future perspective

With recent breakthroughs in artificial intelligence, computer-aided diagnosis (CAD) for upper gastrointestinal endoscopy is gaining increasing attention. Main research focuses in this field include automated identification of dysplasia in Barrett's esophagus and detection of early gastric cancers. By helping endoscopists avoid missing and mischaracterizing neoplastic change in both the esophagus and the stomach, these technologies potentially contribute to solving current limitations of gastroscopy. Currently, optical diagnosis of early-stage dysplasia related to Barrett's esophagus can be precisely achieved only by endoscopists proficient in advanced endoscopic imaging, and the false-negative rate for detecting gastric cancer is approximately 10%. Ideally, these novel technologies should work during real-time gastroscopy to provide on-site decision support for endoscopists regardless of their skill; however, previous studies of these topics remain ex vivo and experimental in design. Therefore, the feasibility, effectiveness, and safety of CAD for upper gastrointestinal endoscopy in clinical practice remain unknown, although a considerable number of pilot studies have been conducted by both engineers and medical doctors with excellent results. This review summarizes current publications relating to CAD for upper gastrointestinal endoscopy from the perspective of endoscopists and aims to indicate what is required for future research and implementation in clinical practice.

Spotting malignancies from gastric endoscopic images using deep learning

BACKGROUND

Gastric cancer is a common kind of malignancies, with yearly occurrences exceeding one million worldwide in 2017. Typically, ulcerous and cancerous tissues develop abnormal morphologies through courses of progression. Endoscopy is a routinely adopted means for examination of gastrointestinal tract for malignancy. Early and timely detection of malignancy closely correlate with good prognosis. Repeated presentation of similar frames from gastrointestinal tract endoscopy often weakens attention for practitioners to result in true patients missed out to incur higher medical cost and unnecessary morbidity. Highly needed is an automatic means for spotting visual abnormality and prompts for attention for medical staff for more thorough examination.

METHODS

We conduct classification of benign ulcer and cancer for gastrointestinal endoscopic color images using deep neural network and transfer-learning approach. Using clinical data gathered from Gil Hospital, we built a dataset comprised of 200 normal, 367 cancer, and 220 ulcer cases, and applied the inception, ResNet, and VGGNet models pretrained on ImageNet. Three classes were defined-normal, benign ulcer, and cancer, and three separate binary classifiers were built-those for normal vs cancer, normal vs ulcer, and cancer vs ulcer for the corresponding classification tasks. For each task, considering inherent randomness entailed in the deep learning process, we performed data partitioning and model building experiments 100 times and averaged the performance values.

RESULTS

Areas under curves of respective receiver operating characteristics were 0.95, 0.97, and 0.85 for the three classifiers. The ResNet showed the highest level of performance. The cases involving normal, i.e., normal vs ulcer and normal vs cancer resulted in accuracies above 90%. The case of ulcer vs cancer classification resulted in a lower accuracy of 77.1%, possibly due to smaller difference in appearance than those cases involving normal.

CONCLUSIONS

The overall level of performance of the proposed method was very promising to encourage applications in clinical environments. Automatic classification using deep learning technique as proposed can be used to complement manual inspection efforts for practitioners to minimize dangers of missed out positives resulting from repetitive sequence of endoscopic frames and weakening attentions.

Application of convolutional neural networks for evaluating Helicobacter pylori infection status on the basis of endoscopic images

BACKGROUND AND AIM

We recently reported the role of artificial intelligence in the diagnosis of Helicobacter pylori (H. pylori) gastritis on the basis of endoscopic images. However, that study included only H. pylori-positive and -negative patients, excluding patients after H. pylori-eradication. In this study, we constructed a convolutional neural network (CNN) and evaluated its ability to ascertain all H. pylori infection statuses.

METHODS

A deep CNN was pre-trained and fine-tuned on a dataset of 98,564 endoscopic images from 5236 patients (742 H. pylori-positive, 3649 -negative, and 845 -eradicated). A separate test data set (23,699 images from 847 patients; 70 positive, 493 negative, and 284 eradicated) was evaluated by the CNN.

RESULTS

The trained CNN outputs a continuous number between 0 and 1 as the probability index for H. pylori infection status per image (Pp, H. pylori-positive; Pn, negative; Pe, eradicated). The most probable (largest number) of the three infectious statuses was selected as the 'CNN diagnosis'. Among 23,699 images, the CNN diagnosed 418 images as positive, 23,034 as negative, and 247 as eradicated. Because of the large number of H. pylori negative findings, the probability of H. pylori-negative was artificially re-defined as Pn -0.9, after which 80% (465/582) of negative diagnoses were accurate, 84% (147/174) eradicated, and 48% (44/91) positive. The time needed to diagnose 23,699 images was 261 seconds.

CONCLUSION

We used a novel algorithm to construct a CNN for diagnosing H. pylori infection status on the basis of endoscopic images very quickly.

ABBREVIATIONS

H. pylori: Helicobacter pylori; CNN: convolutional neural network; AI: artificial intelligence; EGD: esophagogastroduodenoscopies.

Multiple White and Flat Elevated Lesions Observed in the Stomach: A Prospective Study of Clinical Characteristics and Risk Factors

Objective Multiple white and flat elevated lesions (MWFLs) observed in the stomach have only been presented in abstracts at academic conferences over the last decade; therefore, relatively little is known about these lesions. Our aim was to prospectively clarify the clinical characteristics of MWFLs, to identify their risk factors and to retrospectively evaluate the clinical progression of these lesions. Methods A prospective analysis of clinical characteristics and risk factors was conducted in participants who underwent esophagogastroduodenoscopic screening at our hospital. A retrospective analysis of the medical chart of patients identified as having MWFLs was conducted to describe the clinical progression of these lesions. Results The prevalence rate of MWFLs was 10.4% (80/767), with the following risk factors identified on a logistic regression analysis: use of proton pump inhibitors [odds ratio (OR), 3.51; 95% confidence interval (CI), 1.92-6.43], female sex (OR, 1.92; 95% CI, 1.19-3.12) and a 1-year increase in age (OR, 1.05; 95% CI, 1.02-1.08). Among the 70 cases with MWFLs observed over a mean duration of 2.3 years, no progression of MWFLs was detected in 67 cases (96%). Among the 3 remaining cases, progression was mild, with none of the lesions progressing to malignancy. Conclusion The use of proton pump inhibitors (PPIs), female sex, and age are risk factors for MWFLs. We believe that endoscopists should recognize these lesions.

Artificial intelligence diagnosis of Helicobacter pylori infection using blue laser imaging-bright and linked color imaging: a single-center prospective study

Background

Deep learning is a type of artificial intelligence (AI) that imitates the neural network in the brain. We generated an AI to diagnose Helicobacter pylori (H. pylori) infection using blue laser imaging (BLI)-bright and linked color imaging (LCI). The aim of this pilot study was to establish an AI diagnosing system that predicts H. pylori infection status using endoscopic images to improve the accuracy and productivity of endoscopic examination.

Methods

A total of 222 enrolled subjects (105 H. pylori-positive) underwent esophagogastroduodenoscopy and a serum test for H. pylori IgG antibodies. During esophagogastroduodenoscopy, an endoscopist sequentially took 3 still images of the lesser curvature of the stomach using white light imaging (WLI), BLI-bright, and LCI. EG-L580NW endoscopic equipment (FUJIFILM Co., Japan) was used for the study. The specifications of the AI were as follows: operating system, Linux; neural network, GoogLeNet; framework, Caffe; graphic processor unit, Geforce GTX TITAN X (NVIDIA Co., USA).

Results

The area under the curve (AUC) on receiver operating characteristics analysis was 0.66 for WLI. In contrast, the AUCs of BLI-bright and LCI were 0.96 and 0.95, respectively. The AUCs obtained for BLI-bright and LCI were significantly larger than those for WLI (P<0.01).

Conclusions

The results demonstrate that the developed AI has an excellent ability to diagnose H. pylori infection using BLI-bright and LCI. AI technology with image-enhanced endoscopy is likely to become a useful image diagnostic tool.

Training Effect on the Inter-observer Agreement in Endoscopic Diagnosis and Grading of Atrophic Gastritis according to Level of Endoscopic Experience

BACKGROUND

Endoscopic diagnosis of atrophic gastritis can contribute to risk stratification and thereby tailored screening for gastric cancer. We aimed to evaluate the effect of training on inter-observer agreement in diagnosis and grading of endoscopic atrophic gastritis (EAG) according to the level of endoscopists' experience.

METHODS

Twelve endoscopists (six less-experienced and six experienced) participated in this prospective study. The training session consisted of 1) four interventions with two-week intervals, and 2) a follow-up period (two follow-up assessments without feedback). EAG was categorized as C1 to O3 according to the Kimura-Takemoto classification. Kappa statistics were used to calculate inter-observer agreement.

RESULTS

At baseline, kappa indexes were 0.18 in the less-experienced group and 0.32 in the experienced group, respectively. After four interventions with feedback, the kappa index improved in both groups and was sustained during the follow-up period. Overall diagnostic yields of EAG were 43.1% ± 10.7% in pre-intervention and 46.8% ± 5.9% in post-intervention. Variability in the rate of diagnosis of EAG significantly decreased in the less-experienced group (r = 0.04, P = 0.003).

CONCLUSION

Irrespective of experience level, inter-observer agreement for diagnosis and grading of EAG improved after training and remained stable after intervention.

Non-invasive diagnostic tests for Helicobacter pylori infection

BACKGROUND

Helicobacter pylori (H pylori) infection has been implicated in a number of malignancies and non-malignant conditions including peptic ulcers, non-ulcer dyspepsia, recurrent peptic ulcer bleeding, unexplained iron deficiency anaemia, idiopathic thrombocytopaenia purpura, and colorectal adenomas. The confirmatory diagnosis of H pylori is by endoscopic biopsy, followed by histopathological examination using haemotoxylin and eosin (H & E) stain or special stains such as Giemsa stain and Warthin-Starry stain. Special stains are more accurate than H & E stain. There is significant uncertainty about the diagnostic accuracy of non-invasive tests for diagnosis of H pylori.

OBJECTIVES

To compare the diagnostic accuracy of urea breath test, serology, and stool antigen test, used alone or in combination, for diagnosis of H pylori infection in symptomatic and asymptomatic people, so that eradication therapy for H pylori can be started.

SEARCH METHODS

We searched MEDLINE, Embase, the Science Citation Index and the National Institute for Health Research Health Technology Assessment Database on 4 March 2016. We screened references in the included studies to identify additional studies. We also conducted citation searches of relevant studies, most recently on 4 December 2016. We did not restrict studies by language or publication status, or whether data were collected prospectively or retrospectively.

SELECTION CRITERIA

We included diagnostic accuracy studies that evaluated at least one of the index tests (urea breath test using isotopes such as 13C or 14C, serology and stool antigen test) against the reference standard (histopathological examination using H & E stain, special stains or immunohistochemical stain) in people suspected of having H pylori infection.

DATA COLLECTION AND ANALYSIS

Two review authors independently screened the references to identify relevant studies and independently extracted data. We assessed the methodological quality of studies using the QUADAS-2 tool. We performed meta-analysis by using the hierarchical summary receiver operating characteristic (HSROC) model to estimate and compare SROC curves. Where appropriate, we used bivariate or univariate logistic regression models to estimate summary sensitivities and specificities.

MAIN RESULTS

We included 101 studies involving 11,003 participants, of which 5839 participants (53.1%) had H pylori infection. The prevalence of H pylori infection in the studies ranged from 15.2% to 94.7%, with a median prevalence of 53.7% (interquartile range 42.0% to 66.5%). Most of the studies (57%) included participants with dyspepsia and 53 studies excluded participants who recently had proton pump inhibitors or antibiotics.There was at least an unclear risk of bias or unclear applicability concern for each study.Of the 101 studies, 15 compared the accuracy of two index tests and two studies compared the accuracy of three index tests. Thirty-four studies (4242 participants) evaluated serology; 29 studies (2988 participants) evaluated stool antigen test; 34 studies (3139 participants) evaluated urea breath test-13C; 21 studies (1810 participants) evaluated urea breath test-14C; and two studies (127 participants) evaluated urea breath test but did not report the isotope used. The thresholds used to define test positivity and the staining techniques used for histopathological examination (reference standard) varied between studies. Due to sparse data for each threshold reported, it was not possible to identify the best threshold for each test.Using data from 99 studies in an indirect test comparison, there was statistical evidence of a difference in diagnostic accuracy between urea breath test-13C, urea breath test-14C, serology and stool antigen test (P = 0.024). The diagnostic odds ratios for urea breath test-13C, urea breath test-14C, serology, and stool antigen test were 153 (95% confidence interval (CI) 73.7 to 316), 105 (95% CI 74.0 to 150), 47.4 (95% CI 25.5 to 88.1) and 45.1 (95% CI 24.2 to 84.1). The sensitivity (95% CI) estimated at a fixed specificity of 0.90 (median from studies across the four tests), was 0.94 (95% CI 0.89 to 0.97) for urea breath test-13C, 0.92 (95% CI 0.89 to 0.94) for urea breath test-14C, 0.84 (95% CI 0.74 to 0.91) for serology, and 0.83 (95% CI 0.73 to 0.90) for stool antigen test. This implies that on average, given a specificity of 0.90 and prevalence of 53.7% (median specificity and prevalence in the studies), out of 1000 people tested for H pylori infection, there will be 46 false positives (people without H pylori infection who will be diagnosed as having H pylori infection). In this hypothetical cohort, urea breath test-13C, urea breath test-14C, serology, and stool antigen test will give 30 (95% CI 15 to 58), 42 (95% CI 30 to 58), 86 (95% CI 50 to 140), and 89 (95% CI 52 to 146) false negatives respectively (people with H pylori infection for whom the diagnosis of H pylori will be missed).Direct comparisons were based on few head-to-head studies. The ratios of diagnostic odds ratios (DORs) were 0.68 (95% CI 0.12 to 3.70; P = 0.56) for urea breath test-13C versus serology (seven studies), and 0.88 (95% CI 0.14 to 5.56; P = 0.84) for urea breath test-13C versus stool antigen test (seven studies). The 95% CIs of these estimates overlap with those of the ratios of DORs from the indirect comparison. Data were limited or unavailable for meta-analysis of other direct comparisons.

AUTHORS' CONCLUSIONS

In people without a history of gastrectomy and those who have not recently had antibiotics or proton ,pump inhibitors, urea breath tests had high diagnostic accuracy while serology and stool antigen tests were less accurate for diagnosis of Helicobacter pylori infection.This is based on an indirect test comparison (with potential for bias due to confounding), as evidence from direct comparisons was limited or unavailable. The thresholds used for these tests were highly variable and we were unable to identify specific thresholds that might be useful in clinical practice.We need further comparative studies of high methodological quality to obtain more reliable evidence of relative accuracy between the tests. Such studies should be conducted prospectively in a representative spectrum of participants and clearly reported to ensure low risk of bias. Most importantly, studies should prespecify and clearly report thresholds used, and should avoid inappropriate exclusions.

Deep learning analyzes Helicobacter pylori infection by upper gastrointestinal endoscopy images

BACKGROUND AND STUDY AIMS

Helicobacter pylori (HP)-associated chronic gastritis can cause mucosal atrophy and intestinal metaplasia, both of which increase the risk of gastric cancer. The accurate diagnosis of HP infection during routine medical checks is important. We aimed to develop a convolutional neural network (CNN), which is a machine-learning algorithm similar to deep learning, capable of recognizing specific features of gastric endoscopy images. The goal behind developing such a system was to detect HP infection early, thus preventing gastric cancer.

PATIENTS AND METHODS

 For the development of the CNN, we used 179 upper gastrointestinal endoscopy images obtained from 139 patients (65 were HP-positive: ≥ 10 U/mL and 74 were HP-negative: < 3 U/mL on HP IgG antibody assessment). Of the 179 images, 149 were used as training images, and the remaining 30 (15 from HP-negative patients and 15 from HP-positive patients) were set aside to be used as test images. The 149 training images were subjected to data augmentation, which yielded 596 images. We used the CNN to create a learning tool that would recognize HP infection and assessed the decision accuracy of the CNN with the 30 test images by calculating the sensitivity, specificity, and area under the receiver operating characteristic (ROC) curve (AUC).

RESULTS

 The sensitivity and specificity of the CNN for the detection of HP infection were 86.7 % and 86.7 %, respectively, and the AUC was 0.956.

CONCLUSIONS

 CNN-aided diagnosis of HP infection seems feasible and is expected to facilitate and improve diagnosis during health check-ups.

Application of Convolutional Neural Networks in the Diagnosis of Helicobacter pylori Infection Based on Endoscopic Images

BACKGROUND AND AIMS

The role of artificial intelligence in the diagnosis of Helicobacter pylori gastritis based on endoscopic images has not been evaluated. We constructed a convolutional neural network (CNN), and evaluated its ability to diagnose H. pylori infection.

METHODS

A 22-layer, deep CNN was pre-trained and fine-tuned on a dataset of 32,208 images either positive or negative for H. pylori (first CNN). Another CNN was trained using images classified according to 8 anatomical locations (secondary CNN). A separate test data set (11,481 images from 397 patients) was evaluated by the CNN, and 23 endoscopists, independently.

RESULTS

The sensitivity, specificity, accuracy, and diagnostic time were 81.9%, 83.4%, 83.1%, and 198s, respectively, for the first CNN, and 88.9%, 87.4%, 87.7%, and 194s, respectively, for the secondary CNN. These values for the 23 endoscopists were 79.0%, 83.2%, 82.4%, and 230±65min (85.2%, 89.3%, 88.6%, and 253±92min by 6 board-certified endoscopists), respectively. The secondary CNN had a significantly higher accuracy than endoscopists (by 5.3%; 95% CI, 0.3-10.2).

CONCLUSION

H. pylori gastritis could be diagnosed based on endoscopic images using CNN with higher accuracy and in a considerably shorter time compared to manual diagnosis by endoscopists.

Mexican consensus on dyspepsia

Since the publication of the 2007 dyspepsia guidelines of the Asociación Mexicana de Gastroenterología, there have been significant advances in the knowledge of this disease. A systematic search of the literature in PubMed (01/2007 to 06/2016) was carried out to review and update the 2007 guidelines and to provide new evidence-based recommendations. All high-quality articles in Spanish and English were included. Statements were formulated and voted upon using the Delphi method. The level of evidence and strength of recommendation of each statement were established according to the GRADE system. Thirty-one statements were formulated, voted upon, and graded. New definition, classification, epidemiology, and pathophysiology data were provided and include the following information: Endoscopy should be carried out in cases of uninvestigated dyspepsia when there are alarm symptoms or no response to treatment. Gastric and duodenal biopsies can confirm Helicobacter pylori infection and rule out celiac disease, respectively. Establishing a strong doctor-patient relationship, as well as dietary and lifestyle changes, are useful initial measures. H2-blockers, proton-pump inhibitors, prokinetics, and antidepressants are effective pharmacologic therapies. H.pylori eradication may be effective in a subgroup of patients. There is no evidence that complementary and alternative therapies are beneficial, with the exception of Iberogast and rikkunshito, nor is there evidence on the usefulness of prebiotics, probiotics, or psychologic therapies. The new consensus statements on dyspepsia provide guidelines based on up-to-date evidence. A discussion, level of evidence, and strength of recommendation are presented for each statement.

Diagnosis of Helicobacter pylori infection: Current options and developments

Accurate diagnosis of Helicobacter pylori (H. pylori) infection is a crucial part in the effective management of many gastroduodenal diseases. Several invasive and non-invasive diagnostic tests are available for the detection of H. pylori and each test has its usefulness and limitations in different clinical situations. Although none can be considered as a single gold standard in clinical practice, several techniques have been developed to give the more reliable results. Invasive tests are performed via endoscopic biopsy specimens and these tests include histology, culture, rapid urease test as well as molecular methods. Developments of endoscopic equipment also contribute to the real-time diagnosis of H. pylori during endoscopy. Urea breathing test and stool antigen test are most widely used non-invasive tests, whereas serology is useful in screening and epidemiological studies. Molecular methods have been used in variable specimens other than gastric mucosa. More than detection of H. pylori infection, several tests are introduced into the evaluation of virulence factors and antibiotic sensitivity of H. pylori, as well as screening precancerous lesions and gastric cancer. The aim of this article is to review the current options and novel developments of diagnostic tests and their applications in different clinical conditions or for specific purposes.

Comparison of endoscopic based diagnosis with Helicobacter urease test for Helicobacter pylori infection

Background

Helicobacter pylori is an important risk factor for gastritis, peptic ulcers and gastric cancer. The prevalence in developed countries is lower than 40 % but higher than 80 % in some developing countries. It is 75 % in Ghana. The Helicobacter urease test (HUT) is performed at endoscopy and gives an accurate diagnosis. The HUT is not routinely done at our facility and presumption of H. pylori is made based on endoscopic findings and H. pylori eradication prescribed, as the incidence in the general population is presumed high. Is this endoscopic diagnosis sufficient for diagnosing and treating H. pylori? We aimed to assess the feasibility of an endoscopic based H. pylori diagnosis and its accuracy using a HUT as the gold standard in consecutive patients.

Methods

Seventy-six consecutive adult patients with dyspepsia were assessed by upper gastrointestinal endoscopy. A clinical diagnosis of H. pylori or not was made. Biopsy samples were collected for HUT. H. pylori was diagnosed if HUT was positive. The results were then compared.

Results

Median age of patients was 45.0 years. H. pylori prevalence detected by HUT was 51.3 % (95 % CI 40.0–63.0). Sensitivity of endoscopic diagnosis of H. pylori was 71.8 % (95 % CI 55.1–85.0) and specificity was 37.8 % (95 % CI 22.5–55.2). There was no association between clinical findings (73.7 %) and HUT (26.3 %) (OR = 0.80; [95 % CI 0.24–2.64], p = 0.682). There was also no association between endoscopic diagnosis (71.8 %) and HUT (28.2 %), (OR = 1.55; 95 % CI 0.59–4.06, p = 0.373).

Conclusion

Helicobacter pylori infection was not as high as that published in earlier reports. The endoscopic diagnosis alone is not sufficient to make a diagnosis of H. pylori.

Clinicopathologic characteristics of interval gastric cancer in Korea

Background/Aims

Interval gastric cancer (IGC) is defined as cancer that is diagnosed between the time of screening and postscreening esophagogastroduodenoscopy (EGD). Unfortunately, little is known about the characteristics of IGC in Korea, a country with a high incidence of gastric cancer. The aim of this study was to evaluate the clinicopathologic characteristics of IGCs in Korea.

Methods

From January 2006 to July 2011, a total of 81,762 subjects underwent screening EGD at Yonsei University Health Promotion Center, Seoul, Korea. We defined missed cancer as cancer diagnosed within 1 year of screening EGD and latent cancer as cancer diagnosed more than 1 year after EGD.

Results

A total of 16 IGC patients (17 lesions; three missed cancers and 14 latent cancers) were identified, with a mean age of 60.68 years and a mean interval time of 19.64 months. IGCs tended to be undifferentiated (12/17, 70.6%), located in the lower body of the stomach (12/17, 70.6%) and exhibited flat/depressed endoscopic morphology (11/17, 64.7%). The patients with missed cancer were generally younger than the patients with latent cancer (51.3 years vs 62.8 years, p=0.037), and the patients with undifferentiated cancer were significantly younger than those with differentiated cancer (57.0 years vs 68.8 years, p=0.008).

Conclusions

IGCs tended to be undifferentiated, located in the lower body of the stomach, and exhibited flat/depressed endoscopic morphology.

Diagnosis of Helicobacter pylori infection: Current options and developments

Accurate diagnosis of Helicobacter pylori (H. pylori) infection is a crucial part in the effective management of many gastroduodenal diseases. Several invasive and non-invasive diagnostic tests are available for the detection of H. pylori and each test has its usefulness and limitations in different clinical situations. Although none can be considered as a single gold standard in clinical practice, several techniques have been developed to give the more reliable results. Invasive tests are performed via endoscopic biopsy specimens and these tests include histology, culture, rapid urease test as well as molecular methods. Developments of endoscopic equipment also contribute to the real-time diagnosis of H. pylori during endoscopy. Urea breathing test and stool antigen test are most widely used non-invasive tests, whereas serology is useful in screening and epidemiological studies. Molecular methods have been used in variable specimens other than gastric mucosa. More than detection of H. pylori infection, several tests are introduced into the evaluation of virulence factors and antibiotic sensitivity of H. pylori, as well as screening precancerous lesions and gastric cancer. The aim of this article is to review the current options and novel developments of diagnostic tests and their applications in different clinical conditions or for specific purposes.

Diagnosis of Helicobacter pylori infection

The present manuscript focuses on the new information that was published in the field of diagnosis of Helicobacter pylori this past year. While there is little news about the invasive tests, more data are coming concerning the endoscopic features of H. pylori infection. Major efforts were also done to improve molecular detection of the mutations involved in antibiotic resistance. New antibody-based tests (stool antigen test or indirect antibody tests) were also developed.

Helicobacter pylori infection - recent developments in diagnosis

Considering the recommended indications for Helicobacter pylori (H. pylori) eradication therapy and the broad spectrum of available diagnostic methods, a reliable diagnosis is mandatory both before and after eradication therapy. Only highly accurate tests should be used in clinical practice, and the sensitivity and specificity of an adequate test should exceed 90%. The choice of tests should take into account clinical circumstances, the likelihood ratio of positive and negative tests, the cost-effectiveness of the testing strategy and the availability of the tests. This review concerns some of the most recent developments in diagnostic methods of H. pylori infection, namely the contribution of novel endoscopic evaluation methodologies for the diagnosis of H. pylori infection, such as magnifying endoscopy techniques and chromoendoscopy. In addition, the diagnostic contribution of histology and the urea breath test was explored recently in specific clinical settings and patient groups. Recent studies recommend enhancing the number of biopsy fragments for the rapid urease test. Bacterial culture from the gastric biopsy is the gold standard technique, and is recommended for antibiotic susceptibility test. Serology is used for initial screening and the stool antigen test is particularly used when the urea breath test is not available, while molecular methods have gained attention mostly for detecting antibiotic resistance.