A heart artery is not just one pipe carrying one stream of blood. Its smaller branches constantly draw blood away, and a computer model may now estimate where that flow goes—without measuring every drop directly. Ischaemic heart disease is the leading cause of death worldwide, and it results from insufficient blood flow through the heart’s own arteries.

It can lead to chest pain, heart attack, and heart failure. Treatments can restore blood flow, but they should target only narrowings that actually cause ischaemia. The problem is that routine heart-catheter laboratories still lack a direct way to measure that flow, so doctors rely on indirect signs.

So the goal was to build a way to estimate blood flow and connect losses from the main artery to places where important side branches split away. That better represents the pattern found in real coronary arteries. The regional method was checked against direct flow measurements and compared with a simpler method that spreads flow loss along the artery’s entire length.

The core idea is like a city’s water system: flow entering the main pipe is divided between the outlet and side branches, with branch flow inferred from taper around a split. The model lets blood leave the main artery through its side branches, while keeping the basic accounting exact: blood entering equals blood leaving through the end plus blood leaving through the branches.

It estimates how much each branch takes from how the main artery narrows toward the split, using the relationship between the main vessel and its daughter branches. How side-branch losses are assigned changes what the model says is happening inside a narrowed artery.

Spreading losses evenly ignores local narrowing, while assigning them region by region preserves the distinction between a true stenosis and normal narrowing that later recovers. The homogeneous method produced flow estimates that were significantly related to the direct measurements.

That means the calculation captured part of the variation in coronary blood flow. The regional method also had a significant relationship with direct flow measurements, but agreement was not better than with the simpler method.

The more detailed picture did not automatically make the estimates more accurate overall. For the small-vessel circulation, the computer estimates were significantly related to direct measurements, showing a significant correlation between the computer and invasive methods.

The technique identified clinically significant small-vessel disease in seven patients, or thirty-five percent, and achieved eighty percent sensitivity and specificity. Here is the important clue: agreement between the computer and invasive measurements improved when the arteries had a greater burden of disease, measured by pressure changes and by the degree of narrowing seen in artery images.

For the regional method, agreement also related to narrowing judged by visual assessment; more broadly, agreement improved in cases with greater disease burden. Overall, the model can quantify and regionalize blood flow in the major coronary arteries, concentrating flow loss at side branches.

But the regional and homogeneous methods showed no overall difference in performance. The results were weakened because many cases had little narrowing and only a small pressure change. More testing is needed, but the approach may help predict local blood flow, particularly when a narrowing lies close to an important side branch.

The model can estimate and divide blood flow along coronary arteries, especially near a narrowing beside a major branch. It is promising, but still needs more testing before it can guide treatment routinely.