AI tool sharpens blurry heart scans and cuts MRI time by 90%, helping doctors diagnose heart problems faster and more clearly. (CREDIT: Shutterstock)

A blurry heart scan won’t tell you much. But a clear, high-quality image can show exactly how well a heart is working. That’s why scientists at the University of Missouri Columbia recently built a powerful new tool.

Their AI-based model, called TagGen, turns low-quality heart MRI scans into crisp, detailed images—while also cutting scan time by as much as 90%. That’s a game-changer for doctors, and even more so for the people whose lives may depend on the results.

Making Heart Scans Faster and Better

Getting an MRI of the heart often takes between 30 and 90 minutes. During that time, you must lie still and sometimes hold your breath—sometimes for over 20 heartbeats. Even a little movement, like breathing, can blur the image and make it hard for doctors to understand what’s happening inside the heart.

An infographic comparing low resolution MRI images to imaging enhanced by the model TagGen. (CREDIT: Claudia Carver)

Cardiac MRI uses taglines, which are fine grid-like lines placed over the image. These taglines help track muscle movement in the heart. When you breathe or shift during the scan, those taglines blur or fade. That makes it tough to spot weak or damaged areas in the heart. The problem isn’t just that it takes time. The scan might not even give the right information if the images aren’t sharp enough.

This is where TagGen comes in. Built by researchers from the university’s medical and engineering schools, this new AI tool restores the lost details in those blurry scans. It makes the taglines sharper and the images much clearer, even when the scan was fast or a bit shaky.

“If you have a blurry image, you have very few ways to recover the fine details or quality of the image,” explained lead researcher Changyu Sun. “The sharpness reveals very important information for the clinical diagnosis, like if there’s abnormal movement or any dysfunction.”

Related Stories

• Early combination drug therapy slashes heart attack risk
• New light-activated heart tissue revolutionizes heart repair

How the Technology Works

TagGen uses what’s called a cascaded diffusion-based super-resolution model. In simpler terms, it takes a rough, low-resolution image and fills in the fine details using advanced AI. The model works in layers—each layer adds a bit more information and clarity to the image.

The team also connected this model with a technique known as parallel imaging. That allows the machine to gather information faster by using multiple channels at once. Combining these two approaches helps the system both speed up the scan and fix any fuzziness in the image afterward.

This new process doesn’t just save time. It makes it easier for doctors to understand the full picture of heart function. Before this technology, faster scans meant lower image quality. But with TagGen, you can have both speed and clarity.

Illustrations of the framework of conditional denoising diffusion model. (CREDIT: Changyu Sun, et al.)

What It Means for Patients

When doctors scan your heart, they’re looking for very small problems in how your heart beats, pumps, and moves. Even tiny areas that don’t contract properly could be early signs of disease or damage. If the image is blurry, those signs might be missed. With TagGen, the taglines that track heart muscle movement stay sharp, even in quick scans.

Sun pointed out one of the key benefits: “During a heart MRI scan, patients are asked to hold their breath to reduce chest movement from breathing, which helps create clearer images,” he said. “Some scans take more than 20 heartbeats, making it harder for patients to hold their breath.

By using TagGen to maintain the taglines, doctors can see information they would have otherwise missed, and patients only need to hold their breath for three heartbeats. This technology will lead to better diagnoses and improved patient outcomes.”

That’s a huge relief for many people—especially older adults or those with lung or heart conditions—who may find it hard to hold their breath for long periods.

Comparisons of REGAIN and TagGen for super resolution of synthetic low-resolution MR tagging images for a patient with a rate-3.3 acceleration of the two-chamber view at end-diastole and end-systole. (CREDIT: Changyu Sun, et al.)

A Future of Smarter, Faster Imaging

TagGen could change how doctors across the world look at heart health. With faster scans, hospitals could help more patients each day. Lower scanning times also mean less stress, lower costs, and fewer repeat scans due to poor image quality.

Beyond the heart, this technology may help with other parts of the body that move during scans, like lungs or stomach. Anywhere there’s motion, TagGen’s AI could help clean up the picture.

More research is still needed to fully explore what this system can do, but early results show strong promise. By making use of both AI and advanced imaging science, the researchers have shown how technology can reduce cost and discomfort without losing accuracy.

Better scans mean better care. When doctors see clearly, they can act quickly—and that could save lives.

Research findings are available online in the journal Magnetic Resonance in Medicine.