ISIC-DICM-17K
Description
ISIC-DICM-17K (ISIC Dermoscopic Images and Clinical Metadata 17K) is a curated and balanced dataset derived from the International Skin Imaging Collaboration (ISIC) Archive Gallery. It comprises 17,060 dermoscopic images and clinical metadata (8,530 melanoma and 8,530 non-melanoma classes).
For more details, please follow the project’s GitHub repository: https://github.com/mmu-dermatology-research/isic-dicm-17k
This dataset was used in this study and benchmark to explore the effectiveness of multimodal learning for skin lesion classification:
S. Ahammed, X. Cui, W. Lu and M. H. Yap, "Skin Lesion Classification using Dermoscopic Images and Clinical Metadata: Insights from Multimodal Models," 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Nashville, TN, USA, 2025, pp. 222-230, DOI: 10.1109/CVPRW67362.2025.00027
Files
Dataset Details
- Hospital Clínic de Barcelona
- Memorial Sloan Kettering Cancer Center
- ViDIR Group, Department of Dermatology, Medical University of Vienna
- Hospital Italiano de Buenos Aires
- Department of Dermatology, Hospital Clínic de Barcelona
- The University of Queensland Diamantina Institute, The University of Queensland, Dermatology Research Centre
- Sydney Melanoma Diagnostic Center at Royal Prince Alfred Hospital, Pascale Guitera
- Department of Dermatology, Medical University of Vienna
- Sydney Melanoma Diagnostic Center at Royal Prince Alfred Hospital
- Dr. Julien Anriot
- Anonymous
Licenses
This content is free to use, modify, and share for any purpose, with no restrictions or attribution required.
This content is free to use, modify, and share as long as you provide credit to the original creator.
This content is free to use, modify, and share for non-commercial purposes, as long as you provide credit to the original creator.