Part of a three-part series:
- Setting the Direction for Future Research
- Key Trends Driving Radiology Research
- Collaboration Takes Center Stage
The 2026 ACR® Strategic Research Summit held May 1 in Washington, DC, confirmed for me that radiology research is changing — and changing quickly.
Across presentations and discussions, several recurring themes emerged that are reshaping how research is conceived, conducted and pursued.
Real-world data is front and center
During the Summit, nearly every ACR research committee presentation touched on the value of registries and real-world data in some way. Whether focused on breast cancer, liver disease, interventional radiology or molecular imaging, a growing recognition emerged — traditional clinical trials alone can’t answer all the questions we face.
Registries offer a very different solution. Although they’re not usually fundable through traditional sources like the NIH, they’re scalable, they reflect real practice, and they can generate insights faster and often at lower cost.
We also have an opportunity to connect these registries through shared infrastructure. ACR is already investing in tools that make this possible, helping sites contribute data while maintaining ownership and privacy through ACR Connect.
For example, the ACR National Radiology Data Registry enables researchers to address complex scientific questions and produce results applicable across various care settings, geographic locations and populations.
In addition, as a multisite network, ALZ-NET is collecting real-world data that we can use to advance innovative research and improve clinical care. Join the ALZ-NET Interest List.
AI isn’t just coming — it’s here
AI is actively disrupting many industries, including radiology, but the Summit made it clear that more evidence is needed to ensure that AI implementation will effectively aid radiologists and benefit patients.
Across discussions, one common challenge surfaced: AI tools are being adopted widely in practice, but we don’t always have strong real-world evidence showing how they perform over time.
That opens the door for ACR to solve an important problem! With platforms and tools like Assess-AI from the ACR Data Science Institute® (DSI), there’s an opportunity to:
- Monitor AI performance across institutions
- Compare tools in real-world settings
- Provide guidance on responsible use
The message was clear: Adoption needs to be paired with continuous evaluation.
Standardization matters more than ever
Another strong theme was the need for consistency.
Different imaging protocols, varying reporting styles and inconsistent data structures all make it harder to scale research and apply findings broadly. Across committees, there was strong support for:
- Structured reporting frameworks
- Standardized imaging protocols
- Common evaluation metrics
This holds true for AI development and use, where inconsistent inputs can lead to unpredictable outputs.
Toward more personalized care
Risk-based decision-making is an area of high priority in radiology. In breast imaging, for example, there’s growing interest in moving beyond simple measures like breast density and using more dynamic risk models to guide screening.
This reflects a broader move toward personalized screening — matching the right modality and frequency to the individual patient rather than applying one-size-fits-all guidelines.
A shared direction
While each subspecialty has its own priorities, the overlap across discussions was striking. The same ideas kept surfacing — data infrastructure, AI validation, standardization and collaboration.
That alignment is a good sign. It signals that the field is moving in a shared direction — even across very different domains — and creates a strong foundation for coordinated, high-impact research moving forward.
By Alexandra Kerns, Senior Business Development Manager, ACR Center for Research and Innovation™
