Dataset Overview
Definition：Gastroscopy Endoscopy Video Imaging Datasetis provided byChangsha Langhui Information Technology Co., Ltd.The large-scale gastric mirror inspection video-image label database, built by Long Salang Langhui Information Technology Ltd., is designed to record the full range of goggles, oesophagus, stomach sections, cavity doors and retrospectation observations, to maintain the equipment as a source for the full screening video and key originals, to be accompanied by white-ray inner-sight/ narrow-band imaging/mishrink/dyemogenized inner-scopy multi-model data, and to examine the pathology correlation validation.
| Dataset Name | Gastroscopy Endoscopy Video Imaging Dataset |
| Total Data Volume | Total Endoscopy Volume Included: 1,000,000 Cases |
| Imaging Modality | Endoscopy Video (White Light / NBI / Magnification / Chromoendoscopy) |
| Covered Body Parts | Esophagus/ stomach sections/cunt/ten-ten-tice intestines reach range |
| Raw Data | Original device export complete video and key originals, and maintain original resolution, frame, code, colour/enhanced mode |
| Main Protocols | White-ray endoscopy; narrowband/optic enhancement mode; magnified endoscopy; dyed endoscopy; active examination/treatment session |
| Arrival range, clean/visual quality, anatomy, type/size of stove, morphology, enhancement mode, biopsy | |
| Pathology Correlation | Prioritize active screening pathology, cholesterococcal detection, treatment and follow-up |
| Source Institution | Gastroenterology Department / Endoscopy Center of Partner Grade-A Tertiary Hospitals |
| Use Cases | Diagnosis of stomach mirror AI, early detection of stomach cancer, gastrophotos grade, cholesterosomiasis |
Delivery Specifications and Field Requirements
The data set strictly follows the ENDO-GASTRO uniform delivery index requirements to ensure data quality and traceability.
Delivery & Counting：One full stomach mirror examination in one row; one multi-series/multi-view case only
Coverage：Completely records scope insertion, the esophagus, all anatomical regions of the stomach, the reachable range of the pylorus and duodenum, and withdrawal observation
Positives and Composition：Definite lesions / positive examinations as the main body；"Suspected", "considered", "cannot be excluded", "post-treatment changes" and "pathologically confirmed" are retained separately
Raw Data Requirements：Complete the video and key originals, and keep original resolution, frame, code, colour/enhanced mode, time order
Cross-Data Association：Official endoscopy reports must be linked; video frame labels cannot be assumed to be pathologically diagnosed
2026 Frontier AI Research Progress
Domain Review：The gastroscopy system was sensitive to 0.91, specific 0.93 (Meta analysis) in early detection of stomach cancer in 2025-2026, and the GRAIDS system was approved by the MFS of Korea, increasing the sensitivity of new hands after the use of AI from 72.2 per cent to 96.4 per cent.
Diagnosis of Meta for early stomach cancer in white-ray endoscopy
Frontiers in AI, 2026Meta analysis shows that DL algorithms aggregate sensitivity 0.91 (95% CI: 0.82-0.95) and specificity 0.93 (95% CI: 0.87-0.97) in detection of early stomach cancer in white-ray inner mirrors.
AI Auxiliary developmental abnormalities of stomach cancer and intestinal biotest
Scientific Reports, 2026The AAAAM M2 accuracy rate was 92.51 per cent (91.91 per cent sensitive and 96.12 per cent specific) and the IM detection accuracy rate was about 92 per cent.
AIR-GRAIDS Real-time detection system for stomach cancer
2025The GRAIDS system is 94.2% sensitive, and the newer endoscope physician increased his sensitivity from 72.2% to 96.4% after the use of AI, close to the specialist level of 97%.
Increased NBI endoscope early stomach cancer AI diagnosis
2025Develop a magnification NBI AI diagnostic system to assess diagnostic performance using 500 cases to advance early cancer identification.
Diagnosis of gastric mucous acids for enhanced early cancer detection
2025Based on ATP4B expression, the reduction of visualized gastric mucous acid damage is used to increase the sensitivity of early detection of stomach cancer.
Annotation Workflow & Quality Control
Image Acquisition & De-identification
Standardized acquisition workflow: complete data is exported directly from the equipment, and patient identifying information (PHI) is removed before the data is stored。
Initial Annotation (Specialist Physician)
Attending physicians annotate case by case against the standards, including lesion localization, morphological description and disease term determination。
Review (Associate Chief Physician or Above)
Experts with the title of Associate Chief Physician or above review the initial annotation results item by item, correcting erroneous annotations and supplementing missing dimensions。
Consistency Assessment
10% of samples are randomly selected and independently annotated by 3 physicians to calculate Fleiss' Kappa; below 0.Dimensions scored 75 are flagged for rework。
License and Usage Agreement
Academic Research License
For universities and research institutions, supporting academic research related to medical AI。After signing the agreement, a de-identified data subset is provided; the source must be acknowledged：Langhui Technology DataAssetsAPI。
Commercial License
For medical device companies and AI diagnostics companies, supporting medical AI product R&D and medical device registration。Provides full datasets + custom annotation + incremental update services。
Data Compliance Statement
All images come from legally authorized sources; PHI fields are fully de-identified and contain no information that can directly identify an individual, in compliance with the Personal Information Protection Law and the Data Security Law。
Customized services
Supports extended requirements such as adding specific disease types, multimodal annotation expansion, and custom training of AI-assisted diagnostic models。
Quick Facts
AI Frontier Research
The gastroscopy system was sensitive to 0.91, specific 0.93 (Meta analysis) in early detection of stomach cancer in 2025-2026, and the GRAIDS system was approved by the MFS of Korea, increasing the sensitivity of new hands after the use of AI from 72.2 per cent to 96.4 per cent.