A Poisson Flow Model Maps Cortical Folding Patterns

When brain differences are scattered across the cortex, a single thickness score can miss them. That is the problem the paper tackles in juvenile myoclonic epilepsy, where structural abnormalities are described as subtle and spatially distributed. Instead of collapsing the brain’s shape into one local measure, the authors introduce a Poisson flow model built from gradients of the mean curvature field on the cortical surface. The method turns that information into a smooth scalar field by solving a Poisson equation, and the surface gradient of that field becomes a flow representation of folding organization. In plain terms, it offers a way to trace how sulci and gyri fit together across the cortex, rather than judging each spot in isolation. The paper presents this as a principled geometric framework for studying distributed cortical alterations in JME, where conventional morphometric measures such as cortical thickness have limited sensitivity.

A thickness score can miss a brain change when it is spread across the cortex. That matters in juvenile myoclonic epilepsy, or JME. JME often starts in the teen years. The cortex is the brain's wrinkled outer layer. One spot can look fine. The surface pattern can still carry the clue. This approach asks a different question. Can fold lines be read as a flow, not as isolated points? A fold map can keep that spread-out shape. That is what makes this idea worth a closer look.

When thickness misses the map

The model starts with mean curvature. Mean curvature is a number that tells how bent each patch of cortex is. It then feeds that bend map into a Poisson equation. A Poisson equation is a math rule that makes a patchy map smooth. The output is one scalar field. A scalar field gives one value at each point on the surface. The surface gradient then points along the steepest change. That gradient becomes a flow map of sulcal and gyral order. Sulci are the brain's valleys. Gyri are its ridges. The model keeps them linked as one surface pattern. That gives JME a geometric frame for spread-out change. That is the key shift.

How the flow map is built

The method treats folding as geometry, not just shape. It reads the cortex as a surface. Mean curvature marks how much each patch bends. The Poisson step turns those bends into one smooth map. The smooth map is not the end point. Its surface gradient gives the flow field. A gradient shows the steepest uphill direction. That direction traces how folds sit next to each other. The result stays tied to the cortex itself. It does not flatten the brain into a grid. It gives the cortex a route.

Why the flow matters in JME

JME does not look like one clean scar. Its changes are subtle and spread out. That makes thickness maps easy to underread. The flow model keeps the fold layout intact. It asks how sulci and gyri travel together across the surface. That gives a better fit for a distributed disorder. It also turns a flat score into a map with direction. The map stays anchored to shape.

What should it beat next

The cleanest next test is a head-to-head with thickness, gyrification index, the usual fold score, and regional volumes. Those are the usual summaries in cortical studies. The Poisson flow map should show whether fold order adds new signal. It should also show whether it beats old scores on the same cases. If it does, JME will look less like scattered noise. It will look more like a route across the cortex. That would turn a hidden layout into a visible path.