A photographer shoots a runner with a long exposure and gets a blurry silhouette. That's roughly how a standard MRI scanner "sees" moving organs: the heart contracts, the lungs breathe, and the image smears. To avoid this, patients are asked to freeze and hold their breath. But for a child or someone short of breath, that's nearly impossible.
The method called GRASP tackles the problem differently. Instead of collecting data line by line, it takes radial slices, like spokes on a wheel that rotate by a "golden angle" (about 137.5°). This angle is no accident: it fills the circle evenly over time, even if scanning is interrupted. Neural networks then sort the acquired "spokes" by breathing and heartbeat phases. It's as if a photographer could stitch a sharp image from countless split-second frames.
Over ten years, GRASP has gone from a lab idea to clinical machines. Today it helps not only cardiologists but also oncologists: you can watch a tumor take up contrast without making the patient hold still. Newer versions add a third dimension — volumetric imaging. Each heartbeat becomes a 3D movie. And paired with self-supervised algorithms, the scanner learns to predict motion and remove artifacts even during the scan.
So we see the runner again, but now the photographer snaps the shutter thousands of times per second and assembles a perfect portrait of motion. That's how GRASP turns continuous scanning into a clear image. Doctors can peer into a living, breathing body without a single "hold your breath."
The golden angle of 137.5° ensures the most even distribution of "spokes" around a circle over any number of turns — a property first noted by Leonardo da Vinci while studying leaf arrangement.
In the movie Total Recall, the hero undergoes an instant full-body scan — GRASP with volumetric reconstruction brings that sci-fi closer to reality, minus the harmful radiation.