EGID: 5,883 Multi-Label Endoscopic Gastritis Images from Shanghai, 2024

Shanghai's Renji Hospital contributed 5,883 high-quality white-light endoscopic images from 229 patients, collected between March and November 2024. The dataset, created by Zeyu Li, provides multi-dimensional clinical annotations validated by expert endoscopists. It is intended to support AI development and clinical training for gastritis diagnosis and cancer risk assessment.

Use Cases

• Train multi-label classifiers for gastritis based on four annotated clinical dimensions
• Develop AI systems for endoscopic H. pylori infection status detection
• Build models to classify gastric atrophy presence and distribution patterns
• Create educational tools for clinical training in gastritis diagnosis

Strengths

• 5,883 high-quality images from 229 patients provide substantial volume for model training
• Annotations validated by two endoscopists with senior review, achieving excellent inter-rater agreement (Cohen’s Kappa: 0.947-1.000)
• Systematic annotation across four independent clinical dimensions (H. pylori status, atrophy presence, atrophy distribution, gastritis type)

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment
• Data reflects a single hospital's patient population and a specific 2024 time range, which may limit generalizability

Provenance

Source

Retrospectively collected at Renji Hospital, Shanghai, using the Olympus CV-290 system.

Collection Method

Images were systematically annotated and validated through independent assessment by two experienced endoscopists, with consensus review by a senior gastroenterologist for discordant cases.

Time Range

March to November 2024

Freshness

Last updated 2026-05-04 15:47:01; freshness should be verified

Geography

Shanghai, China