Nature Study Shows AI Can Pinpoint Plaque Patterns Behind Angina in Women With Nonobstructive Disease
A Nature study reports that AI-enabled plaque characterization from coronary CT can help explain angina in women who have nonobstructive atherosclerosis, a group that often struggles to get a clear diagnosis. The work suggests imaging AI may do more than automate reads: it may surface disease mechanisms that conventional interpretation misses.
Women with chest pain but without obvious coronary blockage have long occupied a frustrating gray zone in cardiology. This study is notable because it uses AI not merely to classify images, but to extract plaque features that appear linked to symptoms, potentially giving clinicians a more biologically grounded explanation for angina in patients who are often told their tests are "normal."
That matters for both diagnosis and credibility. Nonobstructive atherosclerosis has been historically under-recognized, especially in women, and delayed recognition can lead to repeated testing, persistent symptoms, and undertreatment. If AI-based plaque characterization can identify the specific patterns associated with ischemic symptoms, it could help shift care from reassurance to targeted risk management.
The bigger implication is that cardiovascular AI is moving beyond detection toward phenotyping. Tools that can describe plaque composition, burden, and morphology may eventually help stratify risk more precisely than stenosis-centric workflows, which is important because many adverse events arise from lesions that do not look dramatic on a standard read.
Still, the path to practice will depend on validation across scanner types, populations, and care settings. The most important question is not whether the model performs well in a study, but whether it changes management in a way that improves outcomes and reduces the diagnostic whiplash that many women with angina currently experience.