SIIM-ISIC Melanoma Classification Kaggle Challenge
The Society for Imaging Informatics in Medicine (SIIM) and the International Skin Imaging Collaboration (ISIC) worked together to host a Kaggle Machine Learning Challenge on Melanoma Classification, using the ISIC Archive which contains the largest publicly available collection of quality controlled dermoscopic images of skin lesions from contributing sites in Australia, Austria, Greece, Spain, and the US.
In this competition, participants were asked to develop image analysis tools to enable the automated diagnosis of melanoma using patient-level contextual information, a process more similar to a clinical workflow. Standards-based healthcare APIs will be used to reduce the interoperability barriers to clinical implementation post-competition.
3,314 teams and 4,120 competitors participated in the challenge, which received 102,544 submissions.
The Top 3 winning teams took home a total of $30,000 in prize money.
Winning algorithms are being open sourced to improve diagnostic accuracy of clinical dermatologists.
Top 3 Winning Teams
Note: This competition had 2 levels of prizes:
- Leaderboard Prizes - Awarded on the basis of private leaderboard rank. Only selected submissions were ranked on the private leaderboard.
- Special Prizes - Awarded to the top scoring models using or not using patient-level contextual information.
- With Context - Top-scoring model making use of patient-level contextual information
- Without Context - Top-scoring model without using any patient-level contextual information
The “With Context” Special Prize went to the 1st Place Leaderboard Finisher.
The “Without Context” Special prize went to the 2nd Place Leaderboard Finisher