13805
Blog
Aidoc Staff

How Radiology Partners Evaluates AI Clinical Models

Accuracy, Adoption, Customer Success
Table of Contents

Think you need flawless AI models to find value? Think again. 

When it comes to determining if an AI model for clinical diagnostic support is valuable, the most important factor is its effectiveness in the real world. 

It’s this philosophy that’s helped guide Radiology Partners as they determine which AI models to deploy within their practice. 

Here, Jason Poff, MD, Director of Innovation at Radiology Partners, shares how Radiology Partners evaluates clinical model performance to determine which models to deploy. 

Factor 1: Medical Malpractice Avoidance

When evaluating a model, Dr. Poff looks at diagnoses that come with high medical malpractice liability. For example, in Radiology Partners’ data set, they find that bowel ischemia with free air is extremely important to detect. 

Radiology Partners has found that by deploying an intra-abdominal free gas model they have support finding these cases, which have high clinical value. While these cases aren’t common, false positives are worth finding what he calls “needles in a haystack”.

Factor 2: High Clinical Impact

Certain findings are considered high clinical impact, such as a treatable brain aneurysm or unstable c-spine fracture that might need surgical intervention. 

Even if the model that would detect these findings have limited  accuracy, its deployment may still offer meaningful clinical benefits. This can be seen with things like rib fractures in the outpatient setting because they can be overlooked due to the nature of the clinical histories provided in outpatient settings.

Factor 3: High EDR

As patient imaging needs grow, there are certain findings that wouldn’t be ideal left over the weekend or for a few days. For example, a child who comes in via the outpatient setting and has a fractured ankle. The quicker the diagnosis. The better.

In all of these cases, the model — even if not perfect — is helping patients get better care more quickly. And this is especially important in critical cases with high impact findings.

Interested in learning more?

Watch the full webinar, “AI in Radiology: A New Perspective on Impact and Innovation”. 

Explore the Latest AI Insights, Trends and Research

Follow a manual added link

Webinars

Aidoc experts, customers and industry leaders share the latest in AI benefits and adoption.

Follow a manual added link

eBooks

Explore how clinical AI can transform your health system with insights rooted in real-world experiences.

Follow a manual added link

AI Strategy

Learn how to go beyond the algorithm to develop a scalable AI strategy and implementation plan.

Follow a manual added link

Solutions

Explore how Aidoc can help increase hospital efficiency, improve outcomes and demonstrate ROI.

Aidoc Staff

© Copyright 2025 Aidoc