To accelerate the development and adoption of artificial intelligence (AI) in clinical practice, radiologists need to be empowered to create AI tools at their own institutions, to meet their own patient needs. ACR AI-LAB™ offers radiologists tools designed to help them learn the basics of AI and participate directly in the creation, validation and use of healthcare AI.

Accelerating radiologists contributions to AI

With the freely available vendor-neutral framework that ACR AI-LAB offers, radiologists will learn to use AI tools to meet their local needs.

ACR AI-LAB will allow radiologists to:

  • Learn the basics of AI
  • Explore the use of AI in their practices
  • Collaborate and contribute to the development of AI models for diagnostic imaging
  • Create AI models tailored to their local patient population

Protecting patient data

Current developments in AI to improve patient care are being driven by research happening primarily at institutions with extensive informatics and data science resources — and primarily using single-institution patient data. Although AI algorithms are more effective when trained across a wide and diverse array of clinical data, sharing data outside an institution is difficult due to patient privacy concerns.

With ACR AI-LAB, developers will have access to a broader array of patient data to improve algorithm diversity and diminish algorithm bias as compared to single-institution data sources. This could potentially accelerate the development of AI for diagnostic radiology, generating new solutions for patient care.

ACR AI-LAB opportunities in 2019

An early look at ACR AI-LAB will be available at the 2019 ACR Annual Meeting in Washington, DC, from May 18-22, 2019, and will demonstrate how radiologists can apply their clinical knowledge to generate and improve AI. ACR AI-LAB offers a hands-on approach to educate radiologists on AI and demystify the process as algorithms are developed, trained and validated.

In June radiology residents can participate in a 60-day “Resident Breast Density Challenge” where they will use ACR DSI-hosted data and computing to train and develop an algorithm.

In the fall, an algorithm will be made available through ACR AI-LAB for radiologists to train in their local institutions. ACR members will have access to sample data from publicly available patient datasets along with user-friendly computational tools, and can begin learning to annotate datasets and train AI models. Other algorithms will be made available moving forward.