Related Experiment Video
Updated: Mar 11, 2026
05:33
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
Published on: July 11, 2025
1.3K
Pathology Public Datasets for Artificial Intelligence: A Systematic Review
Aniketh Reddy Chinnachinnanagari1, Shyam Sundar Debsarkar1, V B Surya Prasath2
1Department of Computer Science, University of Cincinnati, Cincinnati, OH, 45221, USA.
Journal of Imaging Informatics in Medicine
|March 10, 2026
Summary
This study reviews 151 public histopathology datasets for artificial intelligence (AI) research. It categorizes datasets and analyzes key ones to aid AI model development in cancer diagnosis and treatment.
Area of Science:
- Digital Pathology
- Computational Histopathology
- Artificial Intelligence in Medicine
Background:
- Histopathology is vital for disease diagnosis, particularly cancer, influencing patient treatment.
- Artificial intelligence (AI) in digital pathology enhances diagnostic speed, accuracy, and scalability.
- Well-structured, annotated histopathology datasets are crucial for advancing AI in pathology.
Purpose of the Study:
- To provide a comprehensive overview of publicly available histopathology datasets for AI and machine learning research.
- To categorize datasets based on various parameters and analyze popular examples.
- To guide researchers in selecting appropriate datasets for developing AI models in computational histopathology.
Main Methods:
- Systematic review of publicly available histopathology datasets.
- Identification and collection of 151 relevant datasets across different tissue types and cancers.
- Categorization of datasets by patient count, organ, staining, magnification, scanner, size, collection method, year, and resolution.
- Analysis of key datasets like CAMELYON, TUPAC, MIDOG, MoNuSeg, and BreakHis.
Main Results:
- A curated list of 151 histopathology datasets suitable for AI research was compiled.
- Datasets were categorized comprehensively, detailing their characteristics and potential applications.
- Analysis highlighted popular datasets and their specific uses in computational histopathology.
- Identified gaps in current dataset availability, particularly for multimodal data integration.
Conclusions:
- This review offers a valuable resource for researchers in computational histopathology, aiding dataset selection for AI model development.
- Standardizing images and addressing data gaps, especially in multimodal data, will enhance AI model generalizability and reproducibility.
- Facilitating collaboration and improving dataset accessibility are key to advancing AI-driven diagnostics in pathology.