Missed diagnosis of early gastric cancer or high-grade intraepithelial neoplasia

Artificial intelligence-assisted diagnosis of early gastric cancer: present practice and future prospects

Gastric cancer (GC) occupies the first few places in the world among tumors in terms of incidence and mortality, causing serious harm to human health, and at the same time, its treatment greatly consumes the health care resources of all countries in the world. The diagnosis of GC is usually based on histopathologic examination, and it is very important to be able to detect and identify cancerous lesions at an early stage, but some endoscopists' lack of diagnostic experience and fatigue at work lead to a certain rate of under diagnosis. The rapid and striking development of Artificial intelligence (AI) has helped to enhance the ability to extract abnormal information from endoscopic images to some extent, and more and more researchers are applying AI technology to the diagnosis of GC. This initiative has not only improved the detection rate of early gastric cancer (EGC), but also significantly improved the survival rate of patients after treatment. This article reviews the results of various AI-assisted diagnoses of EGC in recent years, including the identification of EGC, the determination of differentiation type and invasion depth, and the identification of borders. Although AI has a better application prospect in the early diagnosis of ECG, there are still major challenges, and the prospects and limitations of AI application need to be further discussed.

Endoscopists and endoscopic assistants' qualifications, but not their biopsy rates, improve gastric precancerous lesions detection rate

BACKGROUND

Detecting gastric precancerous lesions (GPLs) is critical for the early diagnosis and treatment of gastric cancer. Endoscopy combined with tissue examination is an important method for detecting GPLs. However, negative biopsy results often increase patients' risks, economic burdens, and lead to additional healthcare costs. Improving the detection rate of GPLs and reducing the rate of negative biopsies is currently a key focus in endoscopic quality control.

AIM

To explore the relationships between the endoscopist biopsy rate (EBR), qualifications of endoscopists and endoscopic assistants, and detection rate of GPLs.

METHODS

EBR, endoscopists, and endoscopic assistants were divided into four groups: Low, moderate, high, and very high levels. Multivariable logistic regression analysis was used to analyze the relationships between EBR and the qualifications of endoscopists with respect to the detection rate of positive lesions. Pearson and Spearman correlation analyses were used to evaluate the correlation between EBR, endoscopist or endoscopic assistant qualifications, and the detection rate of positive lesions.

RESULTS

Compared with those in the low EBR group, the odds ratio (OR) values for detecting positive lesions in the moderate, high, and very high EBR groups were 1.12 [95% confidence interval (CI): 1.06-1.19, P < 0.001], 1.22 (95%CI: 1.14-1.31, P < 0.001), and 1.38 (95%CI: 1.29-1.47, P < 0.001), respectively. EBR was positively correlated with the detection rate of gastric precancerous conditions (atrophic gastritis/intestinal metaplasia) (ρ = 0.465, P = 0.004). In contrast, the qualifications of the endoscopists were positively correlated with GPLs detection (ρ = 0.448, P = 0.005). Compared to endoscopists with low qualification levels, those with moderate, high, and very high qualification levels endoscopists demonstrated increased detection rates of GPLs by 13% (OR = 1.13, 95%CI: 0.98-1.31), 20% (OR = 1.20, 95%CI: 1.03-1.39), and 32% (OR = 1.32, 95%CI: 1.15-1.52), respectively. Further analysis revealed that the qualifications of endoscopists were positively correlated with the detection rates of GPLs in the cardia (ρ = 0.350, P = 0.034), angularis (ρ = 0.396, P = 0.015) and gastric body (ρ = 0.453, P = 0.005) but not in the antrum (ρ = 0.292, P = 0.079). Moreover, the experience of endoscopic assistants was positively correlated with the detection rate of precancerous lesions by endoscopists with low or moderate qualifications (ρ = 0.427, P = 0.015).

CONCLUSION

Endoscopists and endoscopic assistants with high/very high qualifications, but not EBR, can improve the detection rate of GPLs. These results provide reliable evidence for the development of gastroscopic quality control indicators.

Early Detection of Malignant Cells in Gastric Lavage via Hexokinase 2 and Single-Cell Sequencing for Gastric Cancer Diagnosis

Objective

Gastric cancer represents a significant global health challenge due to its high prevalence and mortality rates, largely attributed to the limitations of current screening methods, such as endoscopy, which impede early diagnosis. This study presents an innovative method for early detection by identifying exfoliated tumor cells in gastric lavage, aiming to overcome challenges related to patient compliance and the variability in endoscopist expertise.

Methods

Hexokinase 2 (HK2), a metabolic marker, was utilized to identify exfoliated tumor cells with heightened glycolytic activity in gastric lavage fluid. The malignancy of these HK2-positive, high-glycolytic tumor cells was further validated using single-cell sequencing (SCS), specifically through genome-wide copy number variation analysis.

Results

A total of 60 individuals were assessed, including 10 patients with gastric cancer (9 at stage IA and 1 at stage IIA), 26 patients with precancerous lesions, 15 patients with benign gastric conditions, and 9 healthy controls. The HK2 assay demonstrated an 80% diagnostic sensitivity for stages IA and IIA of gastric cancer and a 96% diagnostic specificity in distinguishing benign conditions from healthy controls. Importantly, the assay exhibited 57% sensitivity for cases of severe dysplasia, underscoring its potential for early gastric cancer detection and preventive diagnostics.

Conclusion

The study highlights the feasibility of a novel gastric lavage-based HK2 assay, complemented by SCS for malignancy confirmation, as a highly accurate method for the early detection of gastric cancer. This approach offers a promising alternative to traditional gastroscopy, particularly for early-stage disease, potentially enhancing detection rates and improving patient outcomes.

Artificial Intelligence as a Surrogate for Inspection Time to Assess Completeness in Esophagogastroduodenoscopy: A Prospective, Randomized, Noninferiority Study

INTRODUCTION

The completeness of esophagogastroduodenoscopy (EGD) is a prerequisite for detecting lesions. This study aims to explore whether the quality of complete examinations assisted by artificial intelligence (AI) would be comparable with those conducted within the guideline-recommended inspection time.

METHODS

Patients referred for diagnostic, screening, or surveillance EGD were enrolled at Renmin Hospital of Wuhan University. Patients were randomly assigned to 2 groups in a 1:1 ratio. In the AI-assisted group, endoscopists completed observation of the entire upper gastrointestinal tract with AI assistance. In the control group, endoscopists were instructed to spend no less than 7 minutes on each procedure. The primary outcome was the detection rate of neoplastic lesions. Noninferiority was confirmed when the lower bound of the 95% confidence interval (CI) was greater than the margin of -1.5%.

RESULTS

A total of 1,723 patients were prospectively enrolled between July 3, 2023, and April 7, 2024. Seven hundred ninety-six and 763 patients in the AI-assisted and control groups were included in the final analysis, respectively. The detection rates of neoplastic lesions in the AI-assisted and control groups were 3.14% and 2.36%, respectively, resulting in an absolute proportion difference of 0.78% (95% CI -0.58% to 2.14%; odds ratio 1.342 [95% CI 0.726-2.480]). The median inspection time was reduced by 1.5 minutes in the AI-assisted group (6.18 [2.87] vs 7.70 [1.90], P < 0.001).

DISCUSSION

Inspection time of complete EGD can be significantly shortened by AI without compromising its quality. These findings provide crucial evidence to support that AI-assisted procedural completeness serves as an objective and effective quality indicator for EGD.

A prospective and comparative study on improving the diagnostic accuracy of early gastric cancer based on deep convolutional neural network real-time diagnosis system (with video)

BACKGROUND

Endoscopic diagnosis of early gastric cancer (EGC) is a challenge. It is not clear whether deep convolutional neural network (DCNN) model could improve the endoscopists' diagnostic performance.

METHODS

We established a DCNN-assisted system and found that accuracy of diagnosis is higher than endoscopists. We prospectively collected an independent image test set of 1289 images and a video test set of 130 patients from three endoscopic centers to compare the diagnostic efficacy of 12 endoscopists before and after DCNN model assistance. Accuracy, sensitivity, specificity, time, and AUC were the main indicators for comparison.

RESULTS

The DCNN model discriminated EGC from the control group (including ulcers and chronic gastritis) with an AUC of 0.917, a sensitivity of 93.38% (95% CI 91.09-95.12%), and a specificity of 90.07% (95% CI 87.59-92.10%) in the image dataset. The video test dataset have an AUC of 0.930, a sensitivity of 96.92% (95% CI 88.83-99.78%), and a specificity of 89.23% (95% CI 79.11-94.98%). The diagnostic performance of novice endoscopists was comparable to those of expert endoscopists with the DCNN model's assistance (accuracy: 95.22 vs. 96.16%) in image test dataset. In the video test, the novice endoscopists, accuracy after DCNN assistance was also improved from 79.36 to 86.41%, and from 86.28 to 91.03% for expert endoscopists. The mean pairwise kappa value of endoscopists was increased significantly with the DCNN model's assistance (0.705-0.753 vs.0.767-0.890) in image testing, and (0.657-0.793 vs. 0.738-0.905) in video testing. The diagnostic duration reduced considerably with the assistance of the DCNN model from 7.09 ± 0.6 s to 5.05 ± 0.55 s in image test, and from 2392.17 ± 7.77 s to2378.34 ± 23.51 s in video test.

CONCLUSION

We developed a DCNN-assisted diagnostic system. And the system can improve the diagnostic performance of endoscopists and help novice endoscopists achieve diagnostic accuracy comparable to that of expert endoscopists.

Evaluation of an artificial intelligence-based system for real-time high-quality photodocumentation during esophagogastroduodenoscopy

Complete and high-quality photodocumentation in esophagoduodenogastroscopy (EGD) is essential for accurately diagnosing upper gastrointestinal diseases by reducing blind spot rates. Automated Photodocumentation Task (APT), an artificial intelligence-based system for real-time photodocumentation during EGD, was developed to assist endoscopists in focusing more on the observation rather than repetitive capturing tasks. This study aimed to evaluate the completeness and quality of APT's photodocumentation compared to endoscopists. The dataset comprised 37 EGD videos recorded at Seoul National University Hospital between March and June 2023. Virtual endoscopy was conducted by seven endoscopists and APT, capturing 11 anatomical landmarks from the videos. The primary endpoints were the completeness of capturing landmarks and the quality of the images. APT achieved an average accuracy of 98.16% in capturing landmarks. Compared to that of endoscopists, APT demonstrated similar completeness in photodocumentation (87.72% vs. 85.75%, P = .0.258), and the combined photodocumentation of endoscopists and APT reached higher completeness (91.89% vs. 85.75%, P < .0.001). APT captured images with higher mean opinion scores than those of endoscopists (3.88 vs. 3.41, P < .0.001). In conclusion, APT provides clear, high-quality endoscopic images while minimizing blind spots during EGD in real-time.

Gastric Cardiac Carcinoma: Recent Progress in Clinicopathology, Prognosis, and Early Diagnosis

Gastric cardiac carcinoma (GCC), also known as gastroesophageal junction (GEJ) carcinoma, is a slow-growing fatal cancer that arises in gastric cardiac mucosa in a region of about 2 cm above and 3 cm below the GEJ line. This carcinoma shows clinicopathologic and genomic features similar, but not identical, to gastric noncardiac carcinoma (GNCC). In contrast, GCC is much more complicated than esophageal adenocarcinoma (EA) in clinicopathology, genomics, and prognosis. GCC is heterogeneous geographically, accounting for 20%-50% of all gastric carcinomas in endemic regions in China. Compared with EA, GCC shows a much broader histopathologic spectrum and worse prognosis. Although detailed mechanisms of GCC pathogenesis remain elusive, advanced age, Helicobacter pylori infection, and gastroesophageal reflux disease are key risk factors. Intriguingly, goblet cell intestinal metaplasia may not be an essential initial step toward carcinogenesis in all GCC cases. At present, an accurate diagnosis of early GCC with prompt curative resection is the only realistic hope for dramatically improving patient outcomes. The recently developed liquid biopsy technology for serum cell-free DNA is a promising tool for the detection of early GCC, though many challenges remain and an in-depth investigation is required. Given the recent rapid advances in artificial intelligence, endoscopic technology, and a better understanding of endoscopists for subtle mucosal/vascular changes in early GCC, accurate detection of early GCC in a high proportion of cases would be possible.

Optimizing colorectal cancer segmentation with MobileViT-UNet and multi-criteria decision analysis

Colorectal cancer represents a significant health challenge as one of the deadliest forms of malignancy. Manual examination methods are subjective, leading to inconsistent interpretations among different examiners and compromising reliability. Additionally, process is time-consuming and labor-intensive, necessitating the development of computer-aided diagnostic systems. This study investigates the segmentation of colorectal cancer regions of normal tissue, polyps, high-grade intraepithelial neoplasia, low-grade intraepithelial neoplasia, adenocarcinoma, and serrated Adenoma, using proposed segmentation models: VGG16-UNet, ResNet50-UNet, MobileNet-UNet, and MobileViT-UNet. This is the first study to integrate MobileViT as a UNet encoder. Each model was trained with two distinct loss functions, binary cross-entropy and dice loss, and evaluated using metrics including Dice ratio, Jaccard index, precision, and recall. The MobileViT-UNet+Dice loss emerged as the leading model in colorectal histopathology segmentation, consistently achieving high scores across all evaluation metrics. Specifically, it achieved a Dice ratio of 0.944 ± 0.030 and a Jaccard index of 0.897 ± 0.049, with precision at 0.955 ± 0.046 and Recall at 0.939 ± 0.038 across all classes. To further obtain the best performing model, we employed multi-criteria decision analysis (MCDA) using the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS). This analysis revealed that the MobileViT-UNet+Dice model achieved the highest TOPSIS scores of 1, thereby attaining the highest ranking among all models. Our comparative analysis includes benchmarking with existing works, the results highlight that our best-performing model (MobileViT-UNet+Dice) significantly outperforms existing models, showcasing its potential to enhance the accuracy and efficiency of colorectal cancer segmentation.

Engineered low-pathogenic Helicobacter pylori as orally tumor immunomodulators for the stimulation of systemic immune response

Gastric cancer constitutes a malignant neoplasm characterized by heightened invasiveness, posing significant global health threat. Inspired by the analysis that gastric cancer patients with Helicobacter pylori (H. pylori) infection have higher overall survival, whether H. pylori can be used as therapeutics agent and oral drug delivery system for gastric cancer. Hence, we constructed engineered H. pylori for gastric cancer treatment. A type Ⅱ H. pylori with low pathogenicity, were conjugated with photosensitizer to develop the engineered living bacteria NIR-triggered system (Hp-Ce6). Hp-Ce6 could maintain activity in stomach acid, quickly infiltrate through mucus layer and finally migrate to tumor region owing to the cell morphology and urease of H. pylori. H. pylori, accumulated in the tumor site, severed as vaccine to activate cGAS-STING pathway, and synergistically remodel the macrophages phenotype. Upon irradiation within stomach, Hp-Ce6 directly destroyed tumor cells via photodynamic effect inherited by Ce6, companied by inducing immunogenic tumor cell death. Additionally, Hp-Ce6 exhibited excellent biosafety with fecal elimination and minimal blood absorption. This work explores the feasibility and availability of H. pylori-based oral delivery platforms for gastric tumor and further provides enlightening strategy to utilize H. pylori invariably presented in the stomach as in-situ immunomodulator to enhance antitumor efficacy.

A Systematic Review of Cost-Effectiveness Studies on Gastric Cancer Screening

Gastric cancer (GC) poses notable economic and health burdens in settings where the incidence of disease is prevalent. Some countries have established early screening and treatment programs to address these challenges. The objectives of this systematic review were to summarize the cost-effectiveness of gastric cancer screening presented in the literature and to identify the critical factors that influence the cost-effectiveness of screening. This systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines. Economic evaluation studies of gastric cancer screening were reviewed from SCOPUS and PubMed. The Consolidated Health Economic Evaluation Reporting Standards 2022 (CHEERS 2022) was used to assess the quality of reporting presented in the selected articles. Only primary economic evaluation studies addressing the cost-effectiveness, cost-utility, and cost-benefit of gastric cancer screening were selected. Two reviewers scrutinized the selected articles (title, abstract, and full text) to determine suitability for the systematic review based on inclusion and exclusion criteria. Authors' consensus was relied on where disagreements arose. The main outcome measures of concern in the systematic review were cost, effectiveness (as measured by either quality-adjusted life years (QALY) or life-years saved (LYS)), and incremental cost-effectiveness ratio (ICER) of screening versus either no screening or an alternative screening method. Thirty-one studies were selected for the final review. These studies investigated the cost-effectiveness of GC screening based on either primary, secondary, or a combination of primary and secondary interventions. The main primary intervention was Helicobacter pylori (Hp) screening with eradication, while the main secondary intervention was endoscopic screening. Cost-effectiveness was evaluated against no screening or screening using an alternative method in both observational and model-based studies. Screening was mainly cost-effective in Asian countries or their diasporas where the prevalence of GC was high. GC screening was generally not cost-effective among Western countries. GC screening can be cost-effective, but cost-effectiveness is dependent on context-specific factors, including geographical location, the prevalence of GC in the local population, and the screening tool adopted. However, there is benefit in targeting high-risk population groups in Asian countries and their diaspora for GC screening.

Computer-aided colorectal cancer diagnosis: AI-driven image segmentation and classification

Colorectal cancer is an enormous health concern since it is among the most lethal types of malignancy. The manual examination has its limitations, including subjectivity and data overload. To overcome these challenges, computer-aided diagnostic systems focusing on image segmentation and abnormality classification have been developed. This study presents a two-stage approach for the automatic detection of five types of colorectal abnormalities in addition to a control group: polyp, low-grade intraepithelial neoplasia, high-grade intraepithelial neoplasia, serrated adenoma, adenocarcinoma. In the first stage, UNet3+ was used for image segmentation to locate the anomalies, while in the second stage, the Cross-Attention Multi-Scale Vision Transformer deep learning model was used to predict the type of anomaly after highlighting the anomaly on the raw images. In anomaly segmentation, UNet3+ achieved values of 0.9872, 0.9422, 0.9832, and 0.9560 for Dice Coefficient, Jaccard Index, Sensitivity, Specificity respectively. In anomaly detection, the Cross-Attention Multi-Scale Vision Transformer model attained a classification performance of 0.9340, 0.9037, 0.9446, 0.8723, 0.9102, 0.9849 for accuracy, F1 score, precision, recall, Matthews correlation coefficient, and specificity, respectively. The proposed approach proves its capacity to alleviate the overwhelm of pathologists and enhance the accuracy of colorectal cancer diagnosis by achieving high performance in both the identification of anomalies and the segmentation of regions.

Analysis of the detection rate and clinical characteristics of early gastric cancer by painless gastroscopy and ordinary gastroscopy

OBJECTIVE

To investigate the difference of early gastric cancer (EGC) detection rate and endoscopic characteristics between painless and ordinary electronic gastroscopy, and summarize the clinical data of gastric cancer (GC) patients.

METHODS

Clinical data of 72,000 patients who underwent gastroscopy in the First People Hospital of Huzhou (Zhejiang, China) from January 2016 to December 2021 were retrospectively analyzed. The patients were divided into painless gastroscopy group (observation group, 36,000 cases) and ordinary gastroscopy group (control group, 36,000 cases) according to the examination methods. The detection rate of EGC between the 2 groups and the endoscopic characteristics of EGC lesions between the 2 groups were compared, and the clinical data of GC were summarized.

RESULTS

Painless gastroscopy is safer than ordinary gastroscopy. The detection rate of GC and EGC in the observation group was significantly higher than that in the control group (P < .05); the difference between the 2 groups in the detection rate of advanced GC was not statistically significant. The average length of EGC lesions in the observation group was significantly shorter than that in the control group (P < .05). The proportion of EGC with lesion length <2.0 cm in the observation group was significantly higher than that in the control group (P < .05). The proportion of EGC lesions with type II morphology, normal or pallor mucosal color, and no rupture in mucosa in the control group were significantly lower than that in the observation group, respectively (P < .05). The proportion of EGC distributed in the cardia, fundus and corpus was higher in the observation group than in the control group (P < .05). The incidence of helicobacter pylori (HP) infection, precancerous diseases, first-degree relatives of GC patients, and risk factors in patients with GC was significantly higher than that in non-GC patients (P < .05), multivariate logistic regression analysis showed that these were independent influencing factors for the occurrence of GC.

CONCLUSION

Painless gastroscopy can effectively improve the screening and diagnostic efficiency of EGC, especially for EGC lesions that are not easy to expose the site, small in size, superficial, without obvious mucosal color change or without mucosal breakage. Therefore, the value of painless gastroscopy in EGC screening is worth further promotion and research.

Diagnostic performance of deep-learning-based virtual chromoendoscopy in gastric neoplasms

BACKGROUNDS

Cycle-consistent generative adversarial network (CycleGAN) is a deep neural network model that performs image-to-image translations. We generated virtual indigo carmine (IC) chromoendoscopy images of gastric neoplasms using CycleGAN and compared their diagnostic performance with that of white light endoscopy (WLE).

METHODS

WLE and IC images of 176 patients with gastric neoplasms who underwent endoscopic resection were obtained. We used 1,633 images (911 WLE and 722 IC) of 146 cases in the training dataset to develop virtual IC images using CycleGAN. The remaining 30 WLE images were translated into 30 virtual IC images using the trained CycleGAN and used for validation. The lesion borders were evaluated by 118 endoscopists from 22 institutions using the 60 paired virtual IC and WLE images. The lesion area concordance rate and successful whole-lesion diagnosis were compared.

RESULTS

The lesion area concordance rate based on the pathological diagnosis in virtual IC was lower than in WLE (44.1% vs. 48.5%, p < 0.01). The successful whole-lesion diagnosis was higher in the virtual IC than in WLE images; however, the difference was insignificant (28.2% vs. 26.4%, p = 0.11). Conversely, subgroup analyses revealed a significantly higher diagnosis in virtual IC than in WLE for depressed morphology (41.9% vs. 36.9%, p = 0.02), differentiated histology (27.6% vs. 24.8%, p = 0.02), smaller lesion size (42.3% vs. 38.3%, p = 0.01), and assessed by expert endoscopists (27.3% vs. 23.6%, p = 0.03).

CONCLUSIONS

The diagnostic ability of virtual IC was higher for some lesions, but not completely superior to that of WLE. Adjustments are required to improve the imaging system's performance.

Endoscopic diagnosis and prevalence of early gastric cancer in India: A prospective study

Objectives

Although countries like Japan and South Korea have implemented nationwide endoscopic screening programs, there is limited evidence on the effectiveness of endoscopy for diagnosing early gastric cancer (EGC) in developing countries such as India. In the present study, we aimed to determine the feasibility of endoscopic detection of EGC from India.

Methods

The data was prospectively collected for all patients ≥40 years who underwent a diagnostic upper endoscopy from April to September 2021. A single endoscopist who performed the endoscopic procedures completed 1-month training in advanced endoscopy in Japan. Following the training, the endoscopist continued to engage in internet-based discussions regarding his cases encountered. Prior to this training, the endoscopist had not detected any EGC cases during his 12-year gastroenterology practice.

Results

A total of 1033 patients were included in the study, with males accounting for 65.4% and a mean age 52 years. The average procedural time was 7.13 ± 4.8 min. A total of 25 patients (2.4%) were found to have GC, including 6 patients (0.6%) with EGC. Two patients had synchronous EGC lesions. All EGC patients were males, with an average age of 66 years. All EGCs were detected in the distal stomach in the presence of Helicobacter pylori infection and severe atrophic gastritis.

Conclusion

Our findings showed that the endoscopic detection of EGC is feasible in India. Optimal training on endoscopic diagnosis of EGC can improve the detection of such lesion. Further studies are warranted to assess the optimization and implementation of an endoscopic screening program for EGC in India.

Comparison of the Prognosis of Upper-Third Gastric Cancer With That of Middle and Lower-Third Gastric Cancer

PURPOSE

Gastric cancer is one of the most common cancers in Korea, and the proportion of upper-third gastric cancers has been steadily increasing over the last two decades. This study aimed to evaluate the effect of tumor location on gastric cancer prognosis.

MATERIALS AND METHODS

We retrospectively reviewed 2,466 patients who underwent gastrectomy for pathologically proven gastric cancer between January 2011 and December 2016. The patients were divided into an upper-third group (U group; n=419, 17.0%) and a middle- and lower-third group (ML group; n=2,047, 83.0%). Clinicopathological characteristics, overall survival (OS), and recurrence-free survival (RFS) after surgery were compared.

RESULTS

The U group had more advanced disease than the ML group and a higher incidence of N3b disease for T3 (12.0% vs. 4.9%, p=0.023) and T4 tumors (33.3% vs. 17.5%, p=0.001). The 5-year RFS rate for stage III disease was marginally lower in the U group than that in the ML group (47.1% vs. 56.7%, p=0.082). The upper third location was an independent prognostic factor for both OS (hazard ratio [HR], 1.350; 95% confidence interval [CI], 1.065-1.711) and RFS (HR, 1.430; 95% CI, 1.080-1.823).

CONCLUSIONS

Upper-third gastric cancer shows extensive node metastasis compared to those located more distally in ≥T3 tumors. The upper third location is an independent prognostic factor for both OS and RFS and may have an adverse impact on RFS, particularly in patients with stage III gastric cancer.

The value of machine learning approaches in the diagnosis of early gastric cancer: a systematic review and meta-analysis

BACKGROUND

The application of machine learning (ML) for identifying early gastric cancer (EGC) has drawn increasing attention. However, there lacks evidence-based support for its specific diagnostic performance. Hence, this systematic review and meta-analysis was implemented to assess the performance of image-based ML in EGC diagnosis.

METHODS

We performed a comprehensive electronic search in PubMed, Embase, Cochrane Library, and Web of Science up to September 25, 2022. QUADAS-2 was selected to judge the risk of bias of included articles. We did the meta-analysis using a bivariant mixed-effect model. Sensitivity analysis and heterogeneity test were performed.

RESULTS

Twenty-one articles were enrolled. The sensitivity (SEN), specificity (SPE), and SROC of ML-based models were 0.91 (95% CI: 0.87-0.94), 0.85 (95% CI: 0.81-0.89), and 0.94 (95% CI: 0.39-1.00) in the training set and 0.90 (95% CI: 0.86-0.93), 0.90 (95% CI: 0.86-0.92), and 0.96 (95% CI: 0.19-1.00) in the validation set. The SEN, SPE, and SROC of EGC diagnosis by non-specialist clinicians were 0.64 (95% CI: 0.56-0.71), 0.84 (95% CI: 0.77-0.89), and 0.80 (95% CI: 0.29-0.97), and those by specialist clinicians were 0.80 (95% CI: 0.74-0.85), 0.88 (95% CI: 0.85-0.91), and 0.91 (95% CI: 0.37-0.99). With the assistance of ML models, the SEN of non-specialist physicians in the diagnosis of EGC was significantly improved (0.76 vs 0.64).

CONCLUSION

ML-based diagnostic models have greater performance in the identification of EGC. The diagnostic accuracy of non-specialist clinicians can be improved to the level of the specialists with the assistance of ML models. The results suggest that ML models can better assist less experienced clinicians in diagnosing EGC under endoscopy and have broad clinical application value.

Multi-omics Combined with Machine Learning Facilitating the Diagnosis of Gastric Cancer

Gastric cancer (GC) is a highly intricate gastrointestinal malignancy. Early detection of gastric cancer forms the cornerstone of precision medicine. Several studies have been conducted to investigate early biomarkers of gastric cancer using genomics, transcriptomics, proteomics, and metabolomics, respectively. However, endogenous substances associated with various omics are concurrently altered during gastric cancer development. Furthermore, environmental exposures and family history can also induce modifications in endogenous substances. Therefore, in this study, we primarily investigated alterations in DNA mutation, DNA methylation, mRNA, lncRNA, miRNA, circRNA, and protein, as well as glucose, amino acid, nucleotide, and lipid metabolism levels in the context of GC development, employing genomics, transcriptomics, proteomics, and metabolomics. Additionally, we elucidate the impact of exposure factors, including HP, EBV, nitrosamines, smoking, alcohol consumption, and family history, on diagnostic biomarkers of gastric cancer. Lastly, we provide a summary of the application of machine learning in integrating multi-omics data. Thus, this review aims to elucidate: i) the biomarkers of gastric cancer related to genomics, transcriptomics, proteomics, and metabolomics; ii) the influence of environmental exposure and family history on multiomics data; iii) the integrated analysis of multi-omics data using machine learning techniques.

Transcatheter arterial chemoembolization combined with apatinib and camrelizumab for unresectable advanced gastric or gastroesophageal junction cancer: a single-arm, single-center, retrospective study

Purpose

This study aims to investigate the efficacy and safety of transcatheter arterial chemoembolization (TACE) combined with Apatinib and Camrelizumab for treating unresectable advanced gastric or gastroesophageal junction (G/GEJ) cancer.

Material and methods

In this study, data of patients with unresectable advanced G/GEJ cancer who received TACE combined with Apatinib and Camrelizumab from August 2018 to December 2021 was evaluated. After TACE, patients were given intravenous Camrelizumab 200mg every three weeks and oral apatinib 250mg/day for treatment. The primary endpoint was overall survival (OS), and the secondary endpoints were objective response rate (ORR), disease control rate (DCR), and adverse events (AEs).

Results

A total of 49 patients were enrolled in this study. The median follow-up time was 14.0 months, and the median OS was 20.0 months (95% CI = 13.6-26.4). Two patients (4.08%) achieved complete remission, 28 patients (57.14%) achieved partial remission, 18 patients (36.73%) had stable disease, and 1 patient (2.04%) had disease progression. The ORR was 61.22%, and the DCR was 97.96%. Multivariate Cox regression analysis indicated that age (HR 4.74, 95% CI = 1.674-13.440, P=0.003) and multiple distant metastases (HR 20.916, 95% CI = 4.094-106.808, P = 0.001) were independent risk factors for OS. Most AEs were classified as grade 1-2, the most common being RCCEP (69.39%). There were 5 cases of grade 3-4 adverse events (10.20%). No patients discontinued or reduced the treatment dose due to AEs, and all patients received symptomatic treatment.

Conclusion

TACE combined with Apatinib and Camrelizumab is a safe and effective therapeutic option for patients with unresectable advanced G/GEJ cancer, which can significantly improve the median OS and ORR of patients. And the adverse events (AEs) are tolerable and manageable.

The Role of Artificial Intelligence in Gastric Cancer: Surgical and Therapeutic Perspectives: A Comprehensive Review

Stomach cancer has a high annual mortality rate worldwide necessitating early detection and accurate treatment. Even experienced specialists can make erroneous judgments based on several factors. Artificial intelligence (AI) technologies are being developed rapidly to assist in this field. Here, we aimed to determine how AI technology is used in gastric cancer diagnosis and analyze how it helps patients and surgeons. Early detection and correct treatment of early gastric cancer (EGC) can greatly increase survival rates. To determine this, it is important to accurately determine the diagnosis and depth of the lesion and the presence or absence of metastasis to the lymph nodes, and suggest an appropriate treatment method. The deep learning algorithm, which has learned gastric lesion endoscopyimages, morphological characteristics, and patient clinical information, detects gastric lesions with high accuracy, sensitivity, and specificity, and predicts morphological characteristics. Through this, AI assists the judgment of specialists to help select the correct treatment method among endoscopic procedures and radical resections and helps to predict the resection margins of lesions. Additionally, AI technology has increased the diagnostic rate of both relatively inexperienced and skilled endoscopic diagnosticians. However, there were limitations in the data used for learning, such as the amount of quantitatively insufficient data, retrospective study design, single-center design, and cases of non-various lesions. Nevertheless, this assisted endoscopic diagnosis technology that incorporates deep learning technology is sufficiently practical and future-oriented and can play an important role in suggesting accurate treatment plans to surgeons for resection of lesions in the treatment of EGC.

Determination of an Appropriate Endoscopic Monitoring Interval for Patients with Gastric Precancerous Conditions in China

OBJECTIVE

Gastric precancerous conditions such as atrophic gastritis (AG) and intestinal metaplasia (IM) are considered independent risk factors for gastric cancer (GC). The suitable endoscopic monitoring interval is unclear when we attempt to prevent GC development. This study investigated the appropriate monitoring interval for AG/IM patients.

METHODS

Totally, 957 AG/IM patients who satisfied the criteria for evaluation between 2010 and 2020 were included in the study. Univariate and multivariate analyses were used to determine the risk factors for progression to high-grade intraepithelial neoplasia (HGIN)/GC in AG/IM patients, and to determine an appropriate endoscopic monitoring scheme.

RESULTS

During follow-up, 28 AG/IM patients developed gastric neoplasia lesions including gastric low-grade intraepithelial neoplasia (LGIN) (0.7%), HGIN (0.9%), and GC (1.3%). Multivariate analysis identified H. pylori infection (P=0.022) and extensive AG/IM lesions (P=0.002) as risk factors for HGIN/GC progression (P=0.025).

CONCLUSION

In our study, HGIN/GC was present in 2.2% of AG/IM patients. In AG/IM patients with extensive lesions, a 1-2-year surveillance interval is recommended for early detection of HIGN/GC in AG/IM patients with extensive lesions.

EBHI: A new Enteroscope Biopsy Histopathological H&E Image Dataset for image classification evaluation

BACKGROUND AND PURPOSE

Colorectal cancer has become the third most common cancer worldwide, accounting for approximately 10% of cancer patients. Early detection of the disease is important for the treatment of colorectal cancer patients. Histopathological examination is the gold standard for screening colorectal cancer. However, the current lack of histopathological image datasets of colorectal cancer, especially enteroscope biopsies, hinders the accurate evaluation of computer-aided diagnosis techniques. Therefore, a multi-category colorectal cancer dataset is needed to test various medical image classification methods to find high classification accuracy and strong robustness.

METHODS

A new publicly available Enteroscope Biopsy Histopathological H&E Image Dataset (EBHI) is published in this paper. To demonstrate the effectiveness of the EBHI dataset, we have utilized several machine learning, convolutional neural networks and novel transformer-based classifiers for experimentation and evaluation, using an image with a magnification of 200×.

RESULTS

Experimental results show that the deep learning method performs well on the EBHI dataset. Classical machine learning methods achieve maximum accuracy of 76.02% and deep learning method achieves a maximum accuracy of 95.37%.

CONCLUSION

To the best of our knowledge, EBHI is the first publicly available colorectal histopathology enteroscope biopsy dataset with four magnifications and five types of images of tumor differentiation stages, totaling 5532 images. We believe that EBHI could attract researchers to explore new classification algorithms for the automated diagnosis of colorectal cancer, which could help physicians and patients in clinical settings.

Feasibility of olfactomedin 4 as a molecular biomarker for early diagnosis of gastric neoplasia after intestinal metaplasia

OBJECTIVES

This study discusses whether olfactomedin 4 (OLFM4) could be used as a sensitive and specific biomarker in the early diagnosis of gastric cancer (GC) after gastric intestinal metaplasia (GIM).

METHODS

An integrative analysis combining data derived from the Gene Expression Omnibus (GEO) and cBioPortal databases was performed to investigate the potential molecular biomarker. Immunohistochemistry and quantitative real-time polymerase chain reactions were used to measure the expression of messenger ribonucleic acid (mRNA) and protein by OLFM4. In combination with the gastroscopic findings and the OLFM4 expression in GIM-GC, a predictive model was established. The receiver operator characteristic curve (ROC) was applied to assess the diagnostic value of the model for GIM-GC.

RESULTS

According to the GEO and cBioPortal databases, OLFM4 was identified as a key gene in the diagnosis of GIM-GC. Higher protein expression of OLFM4 was found in GIM and GIM-GC compared with chronic superficial gastritis (GS) (p < 0.05). The positive expression rate of OLFM4 in paracancerous tissue (GCP) was higher than in GIM (p > 0.05). There was no significant difference between GIM-GC and GCP (p > 0.05). The mRNA expression of OLFM4 was similar to the protein expression, and the positive expression rate was higher in early GIM-GC than in GIM (p < 0.05).

CONCLUSION

Olfactomedin 4 could be used as a biomarker for the early diagnosis of GIM-GC, and the logistic predictive model could be an effective tool for increasing the early diagnostic rate.

EBHI-Seg: A novel enteroscope biopsy histopathological hematoxylin and eosin image dataset for image segmentation tasks

Background and purpose

Colorectal cancer is a common fatal malignancy, the fourth most common cancer in men, and the third most common cancer in women worldwide. Timely detection of cancer in its early stages is essential for treating the disease. Currently, there is a lack of datasets for histopathological image segmentation of colorectal cancer, which often hampers the assessment accuracy when computer technology is used to aid in diagnosis.

Methods

This present study provided a new publicly available Enteroscope Biopsy Histopathological Hematoxylin and Eosin Image Dataset for Image Segmentation Tasks (EBHI-Seg). To demonstrate the validity and extensiveness of EBHI-Seg, the experimental results for EBHI-Seg are evaluated using classical machine learning methods and deep learning methods.

Results

The experimental results showed that deep learning methods had a better image segmentation performance when utilizing EBHI-Seg. The maximum accuracy of the Dice evaluation metric for the classical machine learning method is 0.948, while the Dice evaluation metric for the deep learning method is 0.965.

Conclusion

This publicly available dataset contained 4,456 images of six types of tumor differentiation stages and the corresponding ground truth images. The dataset can provide researchers with new segmentation algorithms for medical diagnosis of colorectal cancer, which can be used in the clinical setting to help doctors and patients. EBHI-Seg is publicly available at: https://figshare.com/articles/dataset/EBHI-SEG/21540159/1.

Diagnostic value of circulating lncRNAs for gastric cancer: A systematic review and meta-analysis

Objective

With the prevalence of next-generation sequencing (NGS) technology, a large number of long non-coding RNAs (lncRNAs) have attracted tremendous attention and have been the topic of extensive research on gastric cancer (GC). It was revealed that lncRNAs not only participate in the transduction of various signaling pathways, thus influencing GC genesis and development, but also have the potential for GC diagnosis. Therefore, we aimed to conduct a meta-analysis of previous studies on GC.

Materials and methods

An electronic search was made before August 2021 on databases including PubMed, Embase, and Web of Science. Relevant articles that compare lncRNA expression in GC patients and healthy controls were summarized. We conducted a meta-analysis with the objective of evaluating the ability of lncRNAs in diagnosing GC.

Results

A total of 40 original research studies including 6,772 participants were discussed in this meta-analysis. The overall sensitivity, specificity, and the area under the curve (AUC) were 0.78 (95% CI: 0.75-0.81), 0.79 (95% CI: 0.74-0.83), and 0.85 (95% CI: 0.81-0.87), respectively. The value of pooled diagnostic odds ratios (DORs) was 13.00 (95% CI: 10.00-17.00).

Conclusions

This meta-analysis revealed that serum or plasma lncRNAs have high sensitivity and specificity, which makes lncRNAs clinically feasible in diagnosing GC. The results from this meta-analysis demonstrated that peripheral blood lncRNAs may become novel noninvasive biomarkers in the foreseeable future. At the same time, it should be noted that a greater number of blood samples and more evidence from rigorous multicenter clinical studies are necessary to justify their applicability as cancer biomarkers.

Recent advances in the progress of immune checkpoint inhibitors in the treatment of advanced gastric cancer: A review

Most patients with advanced gastric cancer were treated with palliative therapy, which had a poor curative effect and a short survival time. In recent years, the clinical research of immune checkpoint inhibitors in advanced gastric cancer has made a breakthrough and has become an important treatment for advanced gastric cancer. The modes of immune checkpoint inhibitors in the treatment of advanced gastric cancer include single drug, combined chemotherapy, radiotherapy, and multiple immune drug combination therapy, among which combination therapy shows better clinical efficacy, and a large number of trials are currently exploring more effective combination therapy programs. In this paper, the new clinical research progress of immune checkpoint inhibitors in the treatment of advanced gastric cancer is reviewed, with an emphasis on combination therapy.

Age-related twin-peak prevalence profiles of H. pylori infection, gastritis, GIN and gastric cancer: Analyses of 70,534 patients with gastroscopic biopsies

Objectives

H. pylori (Hp) infection has been indicated in the pathogenesis of gastric diseases including gastric cancer (GC). This study aimed at exploring the relationships between Hp infection and gastric diseases including GC in a large dataset of routine patients undergoing gastroscopy.

Methods

From November 2007 to December 2017, 70,534 first-time visiting patients aged 18–94 years with gastroscopic biopsies were histologically diagnosed and analyzed. Patients’ data were entered twice in an Excel spreadsheet database and analyzed using the SPSS (version 22.0) software package and statistical significance was defined as P<0.05 for all analyses.

Results

The first interesting observation was age-related twin-peak prevalence profiles (TPPs) for Hp infection, gastritis, and advanced diseases with different time spans (TS) between the first and second occurring peaks. Hp infection and gastritis had TPPs occurring at earlier ages than TPPs of gastric introepithelial neoplasia (GIN) and GC. More patients were clustered at the second occurring TPPs. The time spans (TS) from the first occurring peak of Hp infection to the first occurring peaks of other gastric diseases varied dramatically with 0–5 years for gastritis; 5–15 years for GINs, and 5–20 years for GC, respectively. The number of males with Hp infection and gastric diseases, excluding non-atrophic gastritis (NAG), was more than that of females (P<0.001).

Conclusions

We have first observed age-related twin-peak prevalence profiles for Hp infection, gastritis, GIN, and GC, respectively, among a large population of patients undergoing gastroscopy. The second prevalence peak of GC is at ages of 70–74 years indicating that many GC patients would be missed during screening because the cut-off age for screening is 69 years old in China.

Clinical Benefit and Cost Effectiveness of Risk-Stratified Gastric Cancer Screening Strategies in China: A Modeling Study

BACKGROUND AND OBJECTIVE

A new gastric cancer screening scoring system (NGCS) strategy was recommended for the early gastric cancer (GC) screening process in China. The current study aimed to assess the clinical benefits and the cost effectiveness of the NGCS strategy in GC high-risk areas of China from a societal perspective.

METHODS

A Markov microsimulation model was developed to evaluate 30 alternative screening strategies with varying initiation age, including the NGCS strategy, the modified NGCS strategy, and the endoscopic screening strategy with various screening intervals. The primary outcomes included GC mortality, number of endoscopies, quality-adjusted life-years (QALYs), costs, and incremental cost-effectiveness ratios (ICERs). Cost estimates were reported in 2021 USD (US$) and both costs and benefits were discounted at 5% annually. Deterministic and probabilistic sensitivity analyses were performed to evaluate model uncertainty.

RESULTS

Screening with the NGCS strategy from age 40 years (40-NGCS) reduced the GC incidence by 86.4%, which provided the greatest benefit across strategies. Compared with all strategies, at a willingness-to pay threshold of US$17,922 per QALY, the 40-NGCS strategy was a leading cost-effective strategy, with an ICER of US$15,668 per QALY. Results were robust in univariate and probabilistic sensitivity analyses. The probability of the 40-NGCS strategy being cost effective was 0.863.

CONCLUSIONS

The 40-NGCS strategy was an effective and cost-effective strategy to reduce GC incidence and mortality in China. The findings provide important evidence for decision makers to formulate and optimize targeted approaches for GC prevention and control policies in China.

A deep learning-based model improves diagnosis of early gastric cancer under narrow band imaging endoscopy

BACKGROUND

Diagnosis of early gastric cancer (EGC) under narrow band imaging endoscopy (NBI) is dependent on expertise and skills. We aimed to elucidate whether artificial intelligence (AI) could diagnose EGC under NBI and evaluate the diagnostic assistance of the AI system.

METHODS

In this retrospective diagnostic study, 21,785 NBI images and 20 videos from five centers were divided into a training dataset (13,151 images, 810 patients), an internal validation dataset (7057 images, 283 patients), four external validation datasets (1577 images, 147 patients), and a video validation dataset (20 videos, 20 patients). All the images were labeled manually and used to train an AI system using You look only once v3 (YOLOv3). Next, the diagnostic performance of the AI system and endoscopists were compared and the diagnostic assistance of the AI system was assessed. The accuracy, sensitivity, specificity, and AUC were primary outcomes.

RESULTS

The AI system diagnosed EGCs on validation datasets with AUCs of 0.888-0.951 and diagnosed all the EGCs (100.0%) in video dataset. The AI system achieved better diagnostic performance (accuracy, 93.2%, 95% CI, 90.0-94.9%) than senior (85.9%, 95% CI, 84.2-87.4%) and junior (79.5%, 95% CI, 77.8-81.0%) endoscopists. The AI system significantly enhanced the performance of endoscopists in senior (89.4%, 95% CI, 87.9-90.7%) and junior (84.9%, 95% CI, 83.4-86.3%) endoscopists.

CONCLUSION

The NBI AI system outperformed the endoscopists and exerted potential assistant impact in EGC identification. Prospective validations are needed to evaluate the clinical reinforce of the system in real clinical practice.

Quality indicators in esophagogastroduodenoscopy

Esophagogastroduodenoscopy (EGD) has been used to diagnose a wide variety of upper gastrointestinal diseases. In particular, EGD is used to screen high-risk subjects of gastric cancer. Quality control of EGD is important because the diagnostic rate is examiner-dependent. However, there is still no representative quality indicator that can be uniformly applied in EGD. There has been growing awareness of the importance of quality control in improving EGD performance. Therefore, we aimed to review the available and emerging quality indicators for diagnostic EGD.

Performance status of targeted biopsy alone versus Sydney protocol by non-NBI expert gastroenterologist in gastric intestinal metaplasia diagnosis

Background and study aims  According to a recent guideline, patients with gastric intestinal metaplasia (GIM) should have at least five biopsies performed under the Sydney protocol to evaluate for risk of extensive GIM. However, only narrow-band imaging (NBI)-targeted biopsy may be adequate to diagnose extensive GIM. Patients and methods  A cross-sectional study was conducted between November 2019 and October 2020. Patients with histology-proven GIM were enrolled. All patients underwent standard esophagogastroduodenoscopy performed by a gastroenterology trainee. The performing endoscopists took biopsies from either a suspected GIM area (NBI-targeted biopsy) or randomly (if negative for GIM read by NBI) to complete five areas of the stomach as per the Sydney protocol. The gold standard for GIM diagnosis was pathology read by two gastrointestinal pathologists with unanimous agreement. Results  A total of 95 patients with GIM were enrolled and 50 (52.6%) were men with a mean age of 64 years. Extensive GIM was diagnosed in 43 patients (45.3%). The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of NBI-targeted biopsy vs. the Sydney protocol were 88.4% vs.100 %, 90.3% vs. 90.3%, 88.4% vs. 89.6%, 90.3% vs. 100%, and 89.5% vs. 94.7%, respectively. The number of specimens from NBI-targeted biopsy was significantly lower than that from Sydney protocol (311vs.475, P  < 0.001). Conclusions  Both NBI-targeted biopsy and Sydney protocol by a gastroenterologist who was not an expert in NBI and who has experience with diagnosis of at least 60 cases of GIM provided an NPV higher than 90%. Thus, targeted biopsy alone with NBI, which requires fewer specimens, is an alternative option for extensive GIM diagnosis.

Assessment of Cimetropium Bromide Use for the Detection of Gastric Neoplasms During Esophagogastroduodenoscopy

IMPORTANCE

Esophagogastroduodenoscopy (EGD) is a common procedure used to examine upper gastrointestinal diseases. Although cimetropium bromide and other antispasmodic agents are commonly administered as premedication to inhibit peristalsis during EGD examination, there are few data regarding the benefits of cimetropium bromide for the detection of gastric neoplasms.

OBJECTIVE

To investigate the association between the use of cimetropium bromide as premedication and gastric neoplasm detection rates during EGD examination.

DESIGN, SETTING, AND PARTICIPANTS

This propensity score-matched retrospective cohort study included 67 683 participants who received EGD screening at the Health Promotion Center of Seoul St. Mary's Hospital, The Catholic University of Korea, from January 2, 2010, to June 30, 2017. Data were analyzed from April 1 to December 30, 2021.

EXPOSURES

Participants were divided into 2 groups: those who received cimetropium bromide before EGD examination (intervention group) and those who did not (control group).

MAIN OUTCOMES AND MEASURES

Gastric neoplasm detection rates.

RESULTS

Among 67 683 participants, the mean (SD) age was 48.6 (10.8) years, and 36 517 participants (54.0%) were male; all participants were Asian (a racially homogenous population). Of those, 28 280 participants (41.8%; mean [SD] age, 50.3 [10.6] years; 57.8% male) received cimetropium bromide, and 39 403 participants (58.2%; mean [SD] age, 47.4 [10.8] years; 51.2% male) did not. Propensity score matching based on confounding variables yielded 41 670 matched participants (20 835 pairs). Detected lesions included 52 dysplasias (0.12%), 40 early cancers (0.10%), 7 advanced cancers (0.02%), and 3 lymphomas (0.01%). Gastric neoplasm detection rates were significantly higher in the intervention group (63 participants [0.30%]) vs the control group (39 participants [0.19%]; P = .02). A significant difference in the combined detection rate of dysplasia and early gastric cancer was found between those in the intervention group (57 participants [0.27%]) vs the control group (35 participants [0.17%]; P = .02). For small gastric lesions (<1 cm), those who received cimetropium bromide had higher detection rates (24 participants [0.12%]) than those who did not (11 participants [0.05%]; P = .03). Lesions in the gastric body were detected significantly more often in the intervention group (34 participants [0.16%]) vs the control group (15 participants [0.07%]; P = .007). In multivariate analyses involving all 67 683 participants, the use of cimetropium bromide was more likely to detect gastric neoplasms compared with nonuse (odds ratio, 1.42; 95% CI, 1.04-1.95; P = .03).

CONCLUSIONS AND RELEVANCE

In this study, the use of cimetropium bromide as premedication was significantly associated with increased gastric neoplasm detection rates during EGD screening, and lesions in the gastric body were detected more frequently among those who received cimetropium bromide compared with those who did not. These findings suggest that cimetropium bromide may be considered as premedication before EGD examination among individuals with no contraindications.

Feasibility of standardized procedures of white light gastroscopy for clinical practice: A multicenter study in China

OBJECTIVE

We aimed to establish a standardized procedure for white light gastroscopy (WLG) to screen gastric lesions including early gastric cancer (EGC) in China and to verify its efficacy and feasibility in clinical practice.

METHODS

A standardized WLG procedure for outpatients at nine tertiary hospitals in Beijing was established. Clinical information of the participants and details of the endoscopic procedures were recorded.

RESULTS

A total of 1051 participants were enrolled in a baseline conventional endoscopic survey between March 2014 and December 2015, while 2156 patients were enrolled in the standardized WLG operation from January 2016 to June 2017. The procedure time of the standardized procedure was significantly longer than that of the baseline conventional procedure (P = 0.003). More images were obtained during the standardized procedure compared with the baseline conventional procedure (P < 0.001). The overall detection rate of gastric lesions in the standardized procedure group was significantly higher than that in the baseline procedure group (52.5% vs 38.4%, P < 0.01). The satisfaction scores of both participants and endoscopists in the standardized procedure group were significantly higher than in the baseline procedure group.

CONCLUSIONS

Compared with the conventional procedure, standardized WLG procedure significantly improves the detection rate of gastric lesions as well as the satisfaction score of participants and endoscopists despite its longer procedure time. It is effective and feasible in clinical practice in China for the use of currently available endoscopic equipment.

Artificial Intelligence in Gastric Cancer: Identifying Gastric Cancer Using Endoscopic Images with Convolutional Neural Network

Simple Summary

Gastric cancer (GC) is one of the most newly diagnosed cancers and the fifth leading cause of death globally. Previous studies reported that the detection rate of gastric cancer (EGC) at an earlier stage is low, and the overall false-negative rate with esophagogastroduodenoscopy (EGD) is up to 25.8%, which often leads to inappropriate treatment. Accurate diagnosis of EGC can reduce unnecessary interventions and benefits treatment planning. Convolutional neural network (CNN) models have recently shown promising performance in analyzing medical images, including endoscopy. This study shows that an automated tool based on the CNN model could improve EGC diagnosis and treatment decision.

Abstract

Gastric cancer (GC) is one of the most newly diagnosed cancers and the fifth leading cause of death globally. Identification of early gastric cancer (EGC) can ensure quick treatment and reduce significant mortality. Therefore, we aimed to conduct a systematic review with a meta-analysis of current literature to evaluate the performance of the CNN model in detecting EGC. We conducted a systematic search in the online databases (e.g., PubMed, Embase, and Web of Science) for all relevant original studies on the subject of CNN in EGC published between 1 January 2010, and 26 March 2021. The Quality Assessment of Diagnostic Accuracy Studies-2 was used to assess the risk of bias. Pooled sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, and diagnostic odds ratio were calculated. Moreover, a summary receiver operating characteristic curve (SROC) was plotted. Of the 171 studies retrieved, 15 studies met inclusion criteria. The application of the CNN model in the diagnosis of EGC achieved a SROC of 0.95, with corresponding sensitivity of 0.89 (0.88–0.89), and specificity of 0.89 (0.89–0.90). Pooled sensitivity and specificity for experts endoscopists were 0.77 (0.76–0.78), and 0.92 (0.91–0.93), respectively. However, the overall SROC for the CNN model and expert endoscopists was 0.95 and 0.90. The findings of this comprehensive study show that CNN model exhibited comparable performance to endoscopists in the diagnosis of EGC using digital endoscopy images. Given its scalability, the CNN model could enhance the performance of endoscopists to correctly stratify EGC patients and reduce work load.

Development of a long noncoding RNA BC032469-dependent gold nanoparticle molecular beacon for the detection of gastric cancer cells

Aim: Long noncoding RNA (lncRNA) BC032469-dependent gold nanoparticle molecular beacons (AuNP-MB) were constructed for the detection of gastric cancer cells. Materials & methods: The AuNP-MBs were prepared according to well-established procedures based on the Au-S interaction between the gold lattice and thiol functionalized oligonucleotides. More importantly, the stability and targeting ability of AuNP-MB were verified by a series of in vitro and in vivo experiments. Results: The lncRNA-dependent probes were successfully utilized for AuNP-MB-based intracellular imaging, with fluorescence effectively emitted in GC cells, but not in normal cells. Notably, such fluorescent emission was positively correlated with lncRNA BC032469 expression. Conclusion: The authors developed an effective fluorescent imaging probe for the recognition of gastric cancer cells.

Screening for upper gastrointestinal cancers with magnetically controlled capsule gastroscopy: a feasibility study

BACKGROUND

The medical consortium is an intensive and disease-specific association that integrates tertiary public hospitals and medical examination centers in China. We aimed to evaluate the feasibility of the medical consortium for screening upper gastrointestinal (GI) cancers (MCSC) by magnetically controlled capsule gastroscopy (MCCG).

METHODS

6627 asymptomatic subjects underwent MCCG as part of health check-ups in the MCSC between March and November 2018. Relevant clinical data were collected and analyzed.

RESULTS

The MCSC detected 32 patients with upper GI cancer (0.48 %) confirmed by pathology. The detection rate of early gastric cancer was 16.67 % (4 /24). Gastric polyps, ulcers, and submucosal tumors were found in 15.54 %, 3.76 %, and 3.17 % of subjects, respectively. The whole GI preparation and operation process were well tolerated.

CONCLUSIONS

The MCSC was a feasible model for upper GI cancer screening, especially for asymptomatic subjects. Further prospective studies with better operational quality control are warranted.

Prevalence and characteristics of post-gastroscopy gastric cancer: A retrospective study from an academic medical center

BACKGROUND AND STUDY AIMS

Gastric cancer is diagnosed by endoscopy but false negative rates of up to 10% in the west and 40% in Asia have been reported. In Lebanon, little is known about the rates of post-gastroscopy gastric cancer (PGGC), defined as the proportion of patients diagnosed with gastric cancer with a negative previous examination within 2 years of diagnosis. We aimed to examine the rate of PGGC and its risk factors, clinico-pathologic and endoscopic characteristics at a University medical Center.

PATIENTS AND METHODS

Retrospective analysis of patients with histologically proven gastric malignancy over the last 14 years. Patients with history of upper endoscopy preceding the index diagnostic endoscopy by 6 to 24 months were included.

RESULTS

18,976 patients underwent upper endoscopy and gastric cancer was diagnosed in 323 (1.7%). Of those, only 4 (1.2%) had a preceding endoscopy within 6 to 24 months of diagnosis: 3 adenocarcinoma and one MALT lymphoma. Upon review of the initial endoscopy, a mucosal abnormality had been noted in all 4 patients and biopsies taken in 3 were negative for cancer. The mean time to cancer diagnosis was 8 months (range 6-13 months).

CONCLUSION

A small proportion of gastric carcinomas are missed on endoscopy in this study. Patients with endoscopic evidence of mucosal abnormalities and negative biopsies should undergo repeat examination with multiple biopsies. Proper endoscopic technique, lesion recognition and adoption of performance improvement measures are important to optimize endoscopic practice.

Preferences of first-degree relatives of gastric cancer patients for gastric cancer screening: a discrete choice experiment

BACKGROUND

It is very necessary to implement gastric cancer screening in China to reduce the mortality of gastric cancer, but there are no national screening guidelines and programs. Understanding of individual preferences is conducive to formulating more acceptable screening strategies, and discrete choice experiments can quantify individual preferences. In addition, the first-degree relatives of gastric cancer patients are at high risk for gastric cancer. Compared with those without a family history of gastric cancer, the risk of gastric cancer in first-degree relatives of gastric cancer patients is increased by 60%. Therefore, a discrete choice experiment was carried out to quantitatively analyse the preferences of first-degree relatives of gastric cancer patients for gastric cancer screening to serve as a reference for the development of gastric cancer screening strategies.

METHODS

A questionnaire was designed based on a discrete choice experiment, and 342 first-degree relatives of gastric cancer patients were investigated. In STATA 15.0 software, the data were statistically analysed using a mixed logit model.

RESULTS

The five attributes included in our study had a significant influence on the preferences of first-degree relatives of gastric cancer patients for gastric cancer screening (P < 0.05). Participants most preferred the sensitivity of the screening program to be 95% (coefficient = 1.424, P < 0.01) with a willingness to pay 2501.902 Yuan (95% CI, 738.074-4265.729). In addition, the participants' sex and screening experiences affected their preferences. An increase in sensitivity 35 to 95% had the greatest impact on the participants' willingness to choose a gastric cancer screening program.

CONCLUSION

The formulation of gastric cancer screening strategies should be rooted in people's preferences. The influence of sex differences and screening experiences on the preferences of people undergoing screening should be considered, and screening strategies should be formulated according to local conditions to help them play a greater role.

Gastric Cancer Screening Methods: A Comparative Study of Two Scoring Methods

Objective

To evaluate the Li’s and Japanese scoring methods scoring for screening early gastric cancer in a healthy population.

Methods

During January 2016–December 2018, profiles of the healthy people participated in a physical examination in the first people’s Hospital of Shanghai were collected. A total of 342 volunteers, including 137 males and 205 females ageing 40–74, were enrolled. After recording the basic information, all volunteers were scored using the Japan scoring method and the new gastric cancer screening score (ie, Li’s score). The subjects’ work characteristics (ROC curve) were drawn according to the patient’s endoscopic pathological examination to indicate early gastric cancer, to determine the best cut-off point for the diagnosis of early gastric cancer by Japanese scoring and Li’s scoring, respectively. The sensitivity and specificity of both scoring methods were calculated as well.

Results

The area under the ROC curve of Japanese and Li’s score, in the diagnosis of early gastric cancer, was 0.763 and 0.837, respectively. Japanese and Li’s score ≥14 were considered as the best cut-off point. The sensitivity and specificity of Li’s scoring were 63.60% and 91.10%, respectively. The sensitivity and specificity of the Japanese score were 54.50% and 87.50%, respectively. The area under the ROC curve in Li’s scoring is more significant than that in Japanese scoring, and there was a substantial difference in the two methods (P<0.05).

Conclusion

Both Li’s scoring and Japanese scoring have shown good screening value for early gastric cancer in a healthy population, but Li’s scoring is more sensitive/specific than Japanese scoring.

LINC02535 promotes cell growth in poorly differentiated gastric cancer

Background

Abnormal long non‐coding RNA (lncRNA) expression plays important roles in gastric cancer. However, the functions of many lncRNAs in poorly differentiated gastric cancer (PDGC) remain unknown.

Methods

Three sets of paired tissues from patients with PDGC were used, and transcriptome sequencing was performed, followed by the construction and sequencing of a library and mapping of the reads. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways and protein‐protein interaction (PPI) networks were analysed, and canonical pathway significance was calculated among the differentially expressed genes (DEGs; p < 0.05). Gene expression in 30 paired PDGC specimens and four cell lines was validated through quantitative PCR. Cell proliferation, migration, invasion, apoptosis, and wound healing were analysed.

Results

A total of 499 upregulated DEGs and 627 downregulated DEGs were identified between peritumoral and gastric cancer tissues. The proportions of positive and negative correlations between LINC02535 and the DEGs were 98.40% and 92.66%, respectively, while the Spearman's correlation coefficient was greater than 0.5. The PPI network showed that approximately 73.15% of the top five genes were directly correlated with LINC02535 according to the STRING database. Based on KEGG analysis, the functions of LINC02535 target genes were enriched in signalling pathways related to cancer cell growth. Furthermore, cell function studies showed that LINC02535 upregulation contributed to cell proliferation, migration, invasion, and wound healing and that its inhibition facilitated cell apoptosis.

Conclusion

LINC02535 expression was upregulated in PDGC and contributed to cell proliferation, migration, invasion and wound healing, whereas its inhibition in PDGC facilitated cell apoptosis.

Identifying early gastric cancer under magnifying narrow-band images with deep learning: a multicenter study

BACKGROUND AND AIMS

Narrow-band imaging with magnifying endoscopy (ME-NBI) has shown advantages in the diagnosis of early gastric cancer (EGC). However, proficiency in diagnostic algorithms requires substantial expertise and experience. In this study, we aimed to develop a computer-aided diagnostic model for EGM (EGCM) to analyze and assist in the diagnosis of EGC under ME-NBI.

METHODS

A total of 1777 ME-NBI images from 295 cases were collected from 3 centers. These cases were randomly divided into a training cohort (n = 170), an internal test cohort (ITC, n = 73), and an external test cohort (ETC, n = 52). EGCM based on VGG-19 architecture (Visual Geometry Group [VGG], Oxford University, Oxford, UK) with a single fully connected 2-classification layer was developed through fine-tuning and validated on all cohorts. Furthermore, we compared the model with 8 endoscopists with varying experience. Primary comparison measures included accuracy, area under the receiver operating characteristic curve (AUC), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV).

RESULTS

EGCM acquired AUCs of .808 in the ITC and .813 in the ETC. Moreover, EGCM achieved similar predictive performance as the senior endoscopists (accuracy: .770 vs .755, P = .355; sensitivity: .792 vs .767, P = .183; specificity: .745 vs .742, P = .931) but better than the junior endoscopists (accuracy: .770 vs .728, P < .05). After referring to the results of EGCM, the average diagnostic ability of the endoscopists was significantly improved in terms of accuracy, sensitivity, PPV, and NPV (P < .05).

CONCLUSIONS

EGCM exhibited comparable performance with senior endoscopists in the diagnosis of EGC and showed the potential value in aiding and improving the diagnosis of EGC by endoscopists.

Application of Kyoto Classification of Gastritis to gastric cancer screening in a primary hospital

BACKGROUND

Gastric cancer is a common malignant digestive system tumor in China, and its prognosis is closely related to the early diagnosis and treatment. The rate of early gastric cancer is less than 10% in China, which is far less than those in Japan (70%) and Korea (50%). Standardization of gastroscopic diagnosis and finding gastric cancer screening program suitable for China are of great significance.

AIM

To explore the clinical value of Kyoto Classification of Gastritis in gastric cancer screening in primary hospitals.

METHODS

The Kyoto Classification of Gastritisg was used to retrospectively analyze the data of patients who visited Shekou People's Hospital of Shenzhen for digestive symptoms from September 2019 to November 2020 and met the new system for gastric cancer screening requirements. All patients were divided into three groups according to the grading results of the Kyoto Classification of Gastritis: Low-score group (< 2 points), medium-score group (≥ 2 points but < 4 points), and high-score group(≥ 4 points). A comparative analysis was performed on the detection of gastric cancer among the three groups.

RESULTS

A total of 1383 patients were included in this study, including 918 (66.4%) in the low-score group, 290 (20.9%) in the medium-score group, and 175 (12.7%) in the high-score group. The total detection rate of gastric cancer was 3.54% (49/1383). There were significant differences in the detection rates of gastric cancer between any two of the three groups (P < 0.05).

CONCLUSION

The Kyoto Classification of Gastritis can significantly improve the detection rate of early gastric cancer and gastric cancer.

Accuracy of upper endoscopies with random biopsies to identify patients with gastric premalignant lesions who can safely be exempt from surveillance

INTRODUCTION

Guidelines recommend endoscopy with biopsies to stratify patients with gastric premalignant lesions (GPL) to high and low progression risk. High-risk patients are recommended to undergo surveillance. We aimed to assess the accuracy of guideline recommendations to identify low-risk patients, who can safely be discharged from surveillance.

METHODS

This study includes patients with GPL. Patients underwent at least two endoscopies with an interval of 1-6 years. Patients were defined 'low risk' if they fulfilled requirements for discharge, and 'high risk' if they fulfilled requirements for surveillance, according to European guidelines (MAPS-2012, updated MAPS-2019, BSG). Patients defined 'low risk' with progression of disease during follow-up (FU) were considered 'misclassified' as low risk.

RESULTS

334 patients (median age 60 years IQR11; 48.7% male) were included and followed for a median of 48 months. At baseline, 181/334 (54%) patients were defined low risk. Of these, 32.6% were 'misclassified', showing progression of disease during FU. If MAPS-2019 were followed, 169/334 (51%) patients were defined low risk, of which 32.5% were 'misclassified'. If BSG were followed, 174/334 (51%) patients were defined low risk, of which 32.2% were 'misclassified'. Seven patients developed gastric cancer (GC) or dysplasia, four patients were 'misclassified' based on MAPS-2012 and three on MAPS-2019 and BSG. By performing one additional endoscopy 72.9% (95% CI 62.4-83.3) of high-risk patients and all patients who developed GC or dysplasia were identified.

CONCLUSION

One-third of patients that would have been discharged from GC surveillance, appeared to be 'misclassified' as low risk. One additional endoscopy will reduce this risk by 70%.

Development and Validation of an Automatic Image-Recognition Endoscopic Report Generation System: A Multicenter Study

INTRODUCTION:

Conventional gastrointestinal (GI) endoscopy reports written by physicians are time consuming and might have obvious heterogeneity or omissions, impairing the efficiency and multicenter consultation potential. We aimed to develop and validate an image recognition–based structured report generation system (ISRGS) through a multicenter database and to assess its diagnostic performance.

Methods:

First, we developed and evaluated an ISRGS combining real-time video capture, site identification, lesion detection, subcharacteristics analysis, and structured report generation. White light and chromoendoscopy images from patients with GI lesions were eligible for study inclusion. A total of 46,987 images from 9 tertiary hospitals were used to train, validate, and multicenter test (6:2:2). Moreover, 5,699 images were prospectively enrolled from Qilu Hospital of Shandong University to further assess the system in a prospective test set. The primary outcome was the diagnosis performance of GI lesions in multicenter and prospective tests.

Results:

The overall accuracy in identifying early esophageal cancer, early gastric cancer, early colorectal cancer, esophageal varices, reflux esophagitis, Barrett’s esophagus, chronic atrophic gastritis, gastric ulcer, colorectal polyp, and ulcerative colitis was 0.8841 (95% confidence interval, 0.8775–0.8904) and 0.8965 (0.8883–0.9041) in multicenter and prospective tests, respectively. The accuracy of cecum and upper GI site identification were 0.9978 (0.9969–0.9984) and 0.8513 (0.8399–0.8620), respectively. The accuracy of staining discrimination was 0.9489 (0.9396–0.9568). The relative error of size measurement was 4.04% (range 0.75%–7.39%).

DISCUSSION:

ISRGS is a reliable computer-aided endoscopic report generation system that might assist endoscopists working at various hospital levels to generate standardized and accurate endoscopy reports (http://links.lww.com/CTG/A485).

Unraveling the identity of gastric cardiac cancer

The classification of gastric cardiac carcinoma (GCC) is controversial. It is currently grouped with esophageal adenocarcinoma (EAC) as an adenocarcinoma of the gastroesophageal junction (GEJ). Recently, diagnostic criteria for adenocarcinoma in the GEJ were established and GCC was separated from EAC. We viewed published evidence to clarify the GCC entity for better patient management. GCC arises in the cardiac mucosa located from 3 cm below and 2 cm above the GEJ line. Compared with EAC, GCC is more like gastric cancer and affects a higher proportion of female patients, younger patients, those with a lower propensity for reflux disease, a wider histopathologic spectrum, and more complex genomic profiles. Although GCC pathogenesis mechanisms remain unknown, the two-etiology proposal is appealing: in high-risk regions, the Correa pathway with Helicobacter pylori infection, chronic inflammation, low acid and intestinal metaplasia, dysplasia and carcinoma may apply, while in low-risk regions the sequence from reflux toxin-induced mucosal injury and high acid, to intestinal metaplasia, dysplasia and carcinoma may occur. In early GCC a minimal risk of nodal metastasis argues for a role of endoscopic therapy, whereas in advanced GCC, gastric cancer staging rules and treatment strategy appear to be more appropriate than the esophageal cancer staging scheme and therapy for better prognosis stratification and treatment. In this brief review we share recent insights into the epidemiology, histopathology and genetics of GCC and hope that this will stimulate further investigations in order to improve the clinical management of patients with GCC.

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.

An insight into the roles of piRNAs and PIWI proteins in the diagnosis and pathogenesis of oral, esophageal, and gastric cancer

P-Element induced wimpy testis (PIWI)-interacting RNA (piRNA) is a member of the non-coding RNAs family. Four PIWI proteins are found to be expressed in humans. The number of studies focusing on the roles of piRNAs and PIWI proteins in the field of cancer is increasing. Oral, esophageal, and gastric cancers are considered as important causes of death. PIWI proteins and piRNAs are suggested to be involved in the pathogenesis of these diseases. Thus, studying these molecules may be beneficial for finding new therapeutics. Since it is shown that currently used biomarkers for these cancers have low sensitivity and specificity, there is a necessity for identifying novel non-invasive biomarkers which are highly sensitive and specific. This paper will provide an insight into current knowledge about the functions of PIWI proteins and piRNAs in the oral, esophageal, and gastric cancer. We discuss how PIWI proteins and piRNAs can be involved in the pathogenesis of these cancers. Moreover, we review the studies concerning with the roles of PIWI proteins and piRNAs as biomarkers which are used for diagnostic and prognostic purposes.

RGB Pixel Brightness Characteristics of Linked Color Imaging in Early Gastric Cancer: A Pilot Study

Background and Aims

Linked color imaging (LCI) helps screen and diagnose for early gastric cancer by color contrast in different mucosa. RGB (red, green, and blue) pixel brightness quantifies colors, which is relatively objective. Limited studies have combined LCI images with RGB to help screen for early gastric cancer (EGC). We aimed to evaluate the RGB pixel brightness characteristics of EGC and noncancer areas in LCI images.

Methods

We retrospectively reviewed early gastric cancer (EGC) patients and LCI images. All pictures were evaluated by at least two endoscopic physicians. RGB pixel brightness analysis of LCI images was performed in MATLAB software to compare the cancer with noncancer areas. Receiver operating characteristic (ROC) curve was analyzed for sensitivity, specificity, cut-off, and area under the curve (AUC).

Results

Overall, 38 early gastric cancer patients were enrolled with 38 LCI images. Pixel brightness of red, green, and blue in cancer was remarkably higher than those in noncancer areas (190.24 ± 37.10 vs. 160.00 ± 40.35, p < 0.001; 117.96 ± 33.91 vs. 105.33 ± 30.01, p = 0.039; 114.36 ± 34.88 vs. 90.93 ± 30.14, p < 0.001, respectively). Helicobacter plyori (Hp) infection was not relevant to RGB distribution of EGC. Whether the score of Kyoto Classification of Gastritis (KCG) is ≥4 or <4, the pixel brightness of red, green, and blue was not disturbed in both cancer and noncancer (p > 0.05). Receiver operating characteristic (ROC) curve for differentiating cancer from noncancer was calculated. The maximum area under the curve (AUC) was 0.767 in B/G, with a sensitivity of 0.605, a specificity of 0.921, and a cut-off of 0.97.

Conclusions

RGB pixel brightness was useful and more objective in distinguishing early gastric cancer for LCI images.