The American College of Radiology (ACR) Data Science Institute® (DSI) released 90 clinical and non-interpretive use cases for the radiology community in November to more than double the number of freely available use cases on the DSI website to 140.
DSI’s use cases are scenarios describing when the use of AI will be beneficial in a clinical setting. They provide structured data elements for training, testing and monitoring algorithms.
Use cases are an open invitation to those developing AI to create a new AI model or to improve upon an existing one.
New to the 2019 release are 18 non-interpretive use cases that mine clinical data for alternative ways to improve health care medical imaging, such as increased efficiency and improved communication. Non-interpretive use cases in this release predict patient no-shows, update patients on wait times based on scanner and volume features, and provide value in a variety of ways.
There are thousands of potential use cases for narrow AI tasks in each domain of radiology.
DSI’s 15 subspecialty panels of 135 experts developed the newly released use cases over the past year after they were identified as high-value problems solvable by AI.
In the coming months DSI will begin continuously rolling out new use cases as they become available. ACR members interested in becoming involved can request panel membership.