The heart doesn't contract all at once but as a wave: an electrical signal travels through the muscle, and behind it, like ripples spreading on water, a contraction rolls through. In arrhythmia, this wave goes awry: it stalls in one spot, loops in another, and the pump starts working in a disorganized way. Today, the only way to see such a wave from the inside is to insert a catheter with sensors into the heart — not a pleasant procedure.

Engineers proposed a workaround: instead of tracking electricity, track how the heart moves. After all, every electrical impulse immediately turns into a contraction — the muscle tenses, deforms, and pushes against the blood. If you measure these deformations precisely from the outside, you can work out where and when the impulse originated. The task is like figuring out where a stone fell by looking at the ripples on the water.

To do this, they built a neural network with the laws of cardiac mechanics embedded in it: how the muscle stretches, how the fibers are oriented, how pressure transmits through the wall. The network was trained on a simplified model of the left ventricle — the main pumping chamber. It was shown only the "ripples"—the surface motion, and it reconstructed the invisible "wave"—the electrical activation inside.

This approach could give doctors a tool for non-invasive arrhythmia diagnosis: instead of a catheter, just a regular ultrasound or MRI, and then the math takes over. For a specific patient, you could build a digital twin of their heart and test how excitation would spread through it. This is a step toward "digital phenotyping"—describing disease not just by symptoms but by the precise mechanics of the organ.