Radiology Professionals

Radiology Professionals

The starting point for creating an effective AI ecosystem is establishing clinical context and technical requirements.

Defining where AI potential is greatest




We’re working with radiology professionals to define and prioritize TOUCH-AI use cases around clinical scenarios in which AI has the potential to improve patient care.

To consider the broadest range of use cases, we solicit proposals from diagnostic and interventional radiologists, radiation oncologists, medical physicists, radiology and other medical specialty societies, and developers. Then, panels of experts choose the highest value proposals for use case development and use their clinical expertise to develop a narrative for how an AI algorithm would work for that use case — while the ACR provides support.

Partnering to empower AI


Machine learning requires good, annotated data. DSI is making services available to help radiology departments create training and testing datasets around ACR DSI use cases, so that the data is usable by industry.

DSI maintains a public directory of datasets and assists in connecting those with data with developers who want to use the datasets to build algorithms. Ultimately, our goal is to empower machine learning in health care.

Putting patient safety first


Family

For AI to be successful, radiologists must trust that it is safe in a variety of clinical settings and that it will benefit their patients and improve care. Using data from multiple sites, ACR DSI can offer AI algorithms built based on TOUCH-AI use case validation prior to FDA approval. Our use cases also contain data elements so that algorithm performance can be monitored following clinical deployment.

We offer developers a secure method to statistically validate algorithms pre-deploymentand the opportunity to monitor performance data for FDA post-market surveillance and for adaptive learning to improve algorithm performance in the field.

Learn more about how DSI is involved in the regulatory process, or read about how AI is working in clinical settings.


Collaborating with radiology professionals


Many AI developers do not have clinical backgrounds and find it challenging to solve problems in ways that provide value to end users. We put doctors at the center of prioritizing the problems AI can help solve in radiology. This ensures industry focuses from the beginning on problems where AI can have an impact.

If you’re already confident in your understanding and ready to join our group of specialty experts we’ll connect you with opportunities to help guide industry on solving the right problems in the most impactful way. If you are still developing your AI skills, visit our resources for links to AI reading materials, podcasts and video.

Find out how to get involved in the work of ACR DSI.