RSNA organizes data challenges to spur the creation of artificial intelligence (AI) tools for radiology.
Data challenges engage the radiology community to develop datasets useful for training AI systems to perform clinically relevant tasks. Researchers then compete to create applications that perform defined tasks according to specified performance measures. The goal of each challenge is to explore and demonstrate the ways AI can benefit radiology and improve patient care.
These AI data challenges are organized by the RSNA Radiology Informatics Committee. Please direct questions about the AI data challenge program to firstname.lastname@example.org.
How does a data challenge work?
Data challenges take place in two phases.
The training phase
During this phase, participants use a training dataset, which includes the radiologists’ labels, to develop algorithms that replicate their annotations.
The evaluation phase
During this phase, participants apply their algorithms to the testing portion of the dataset, which is provided to them with the annotations withheld.
Results are then compared to the annotations on the testing dataset and an evaluation metric is applied to rate their accuracy and determine the winners.
2019 Intracranial Hemorrhage Detection and Classification Challenge
For the 2019 challenge, we assembled a dataset of over 25,000 brain CTs contributed by four research institutions and worked with volunteers from the American Society of Neuroradiology to label them for the presence of five types of intracranial hemorrhages. We then invited teams of data scientists and radiologists to use this dataset to develop algorithms that can identify and categorize hemorrhages.
The 2019 winners
In 2019, Kaggle recognized the RSNA Intracranial Hemorrhage Detection Challenge as a public good and provided $25,000 in prize money for the winning entries.
Downloadable data sets
Data sets from the 2019 Intracranial Hemorrhage Detection are available for download.
For detailed information about this challenge, visit the Kaggle site.
Past AI challenges
View information about past challenges here: