HyperKvasir - sporedata/researchdesigneR GitHub Wiki
General description
The HyperKvasir dataset is a large, open-access collection of medical images and videos used primarily in the field of gastroenterology, specifically for the study and development of computer-aided diagnostic (CAD) systems for endoscopic procedures. It contains over 110,000 images and 373 videos, making it one of the largest datasets of its kind in gastroenterology.
The HyperKvasir dataset was created to assist in developing machine learning (ML) models for medical image analysis, specifically for tasks related to gastrointestinal (GI) tract diseases. It provides a diverse range of images and videos collected during endoscopy, a procedure used to examine the digestive tract.
Data Types
-
Images: The images are high-quality and sourced from various sections of the GI tract (e.g., esophagus, stomach, colon). These include both normal and abnormal findings in the gastrointestinal tract. Examples of abnormalities include polyps, ulcers, esophagitis, and other GI tract conditions.
-
Videos: The videos capture endoscopic procedures and provide dynamic information that helps in understanding the movement and interaction within the GI tract.
Related publications
Data access
For more information on the HyperKvasir dataset, visit https://osf.io/mh9sj/