Researchers at Cleveland Clinic and Pittsburgh-based Carnegie Mellon University have developed an AI system designed to interpret cardiac MRI scans without manually labeled training data.
The system was trained on more than 13,000 de-identified patient studies from Cleveland Clinic and learned from over 1 million images and hundreds of thousands of motion sequences collected over more than a decade, according to a May 21 news release. Researchers said the model outperformed general-purpose AI systems by more than 35% in some testing scenarios.
Cardiac MRI scans can contain hundreds to thousands of images across multiple views and time points, with interpretation often taking 40 minutes or more for specialists. Instead of relying on manually labeled datasets, researchers trained the model using radiology reports paired with MRI image sequences.
The model demonstrated accuracy rates as high as 99% for certain heart conditions and showed the ability to retrieve similar cases using natural language prompts. Researchers also said the system performed strongly on two separate external datasets, including one from Cleveland Clinic Florida.
The research was published in Nature Communications. Researchers said future work will explore applications including automated report generation and clinical decision support systems.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
