FDA Grants Breakthrough Status to Generative Radiology AI
- •FDA grants breakthrough designation to two generative AI tools for chest X-ray interpretation and reporting.
- •Cognita and Aidoc secured the designations, with Aidoc's First Read targeting four life-threatening clinical findings.
- •New generative models draft full reports, moving beyond traditional AI systems that only flag image anomalies.
The U.S. Food and Drug Administration (FDA) has awarded breakthrough device designation to two generative AI tools designed to interpret chest X-rays and draft radiology reports. This regulatory status indicates that the FDA will prioritize the review process for these systems, which represent a departure from traditional machine learning tools that only flag specific areas of interest for human clinicians. Instead, these generative models analyze entire medical images to draft comprehensive findings for subsequent radiologist review.
In March, the FDA granted the designation to Cognita, a startup established by Stanford University researchers that was acquired by Radiology Partners late last year. Additionally, on June 25, 2026, the radiology AI firm Aidoc announced its own breakthrough designation for a product called First Read. This specific tool is currently authorized to detect and describe four distinct life-threatening clinical findings in patient imaging. These advancements reflect a broader shift where large vision language models (AI systems capable of processing and describing visual data) are increasingly tasked with document generation in clinical workflows, posing new challenges for standard validation and existing regulatory oversight frameworks in healthcare.