AI-Powered Technology Significantly Improves Low-Quality Cardiac MRI Scans

Researchers from the University of Missouri have developed an innovative AI-assisted model called TagGen that revolutionizes cardiac magnetic resonance imaging (MRI) by transforming low-resolution, blurry heart scans into high-quality images. This advancement not only enhances diagnostic accuracy but also drastically reduces the time required for scanning by approximately 90%, making the process more comfortable for patients and more cost-effective for healthcare providers.

Typically, a cardiac MRI scan lasts between 30 to 90 minutes. During this period, patients are asked to hold their breath to minimize chest movement, which helps produce clearer images. However, longer scans—sometimes exceeding 20 heartbeats—pose challenges for patients, especially those with difficulty holding their breath, leading to lower image quality due to movement artifacts.

The TagGen AI model uses a diffusion-based, generative approach to restore the fine details of MRI images that are often missed or blurred in traditional scans. This enhanced clarity reveals vital markers known as taglines, which track muscle movements within the heart. Accurate visualization of these movements allows clinicians to identify areas of abnormal function or damage more effectively. By improving image sharpness and detail, TagGen enables doctors to make quicker and more precise diagnoses.

One of the practical advantages of this technology is its ability to maintain high image quality with significantly shorter scan times. This means patients need to hold their breath only for three heartbeats instead of a longer duration, reducing discomfort and increasing the feasibility of obtaining high-quality images, especially in emergency or routine scans.

Future plans for the research team include refining the TagGen model further and expanding its application to other cardiac MRI scans, computed tomography (CT), and imaging of other organs such as the brain. Lead researcher Dr. Changyu Sun emphasizes that this innovative approach could lead to better disease detection, improved patient outcomes, and more efficient imaging workflows.

This breakthrough has been documented in a paper published in the journal Magnetic Resonance in Medicine, highlighting the potential of AI technology to advance medical imaging and diagnostics.

Source: https://medicalxpress.com/news/2025-06-ai-quality-mri-heart-scans.html