Non-invasive numerical model for the human circulatory system

In the 1966 novel Fantastic Journey, by Isaac Asimov, a miniature submarine and crew travel throughout the human circulatory system in a race against time to save the patient. While the story was purely fictional, recent work in hemodynamics has produced a means to virtually view the entire closed-loop circulatory system.

Faisal Amlani and Niema M. Pahlevan developed a blood flow solver that models major vascular segments, organs, heart chambers, and heart valves.

“Our goal is to develop robust computational tools to serve as a virtual twin for a patient to complement ongoing developments of easy-to-use and possibly non-invasive measurement devices, such as wearable technology, for the general population,” said Pahlevan.

While 3D computational fluid dynamics is currently used for testing and modeling medical devices and therapeutic interventions, the 3D method is limited by boundary condition assumptions that are not always physiologically accurate. Wave propagation and reflection must be evaluated to comprehensively analyze the flow of blood through a closed-loop system.

Because of its high accuracy and efficiency, the researchers use a pseudo-spectral approach based on Fourier continuation. More than 400 blood vessels, including systemic arterial and venous vessels, pulmonary arteries and veins, and heart chambers and valves were modeled. Using the patients’ physiological conditions as input, the simulation proved fast and accurate over thousands of heartbeats.

“The model is robust enough to work for extreme heart rates, vascular conditions, and cardiac conditions,” said author Faisal Amlani. “It has the potential to be used for continuous heart monitoring to predict such events as heart attack and cardiac arrest.”

Source: “A high-order space-time fourier continuation approach for one-dimensional hemodynamics and wave propagation in the entire human circulatory system,” by Faisal Amlani and Niema M. Pahlevan, Physics of Fluids (2025). This article can be accessed at https://doi.org/10.1063/5.0273041 .