P2.18 TOWARDS COMPUTATIONAL DIAGNOSIS OF CORONARY ARTERY DISEASE

Abstract

Flow in the wake of a coronary artery stenosis induces a bruit in the 300–1500 Hz range that can be heard at the chest wall. It has been hypothesised that this sound is caused by turbulence-induced shear waves which travel through the soft tissue of the thorax. This contribution describes a computational mathematical ‘forward solve’ method to simulate these shear waves in a virtual chest of tissue mimicking agarose gel. As the first stage in the development of a noninvasive diagnostic tool we also describe initial results towards the solution of the mathematical inverse problem. That is: to identify the source of the bruit given the surface measured signal. Objectives: To demonstrate proof-of-concept of a novel biotechnology that will use mathematical simulations to provide a non-invasive screening tool for coronary artery disease. Methods: Finite element based forward solvers for soft tissue response (given the source, generate the signal); optimisation-based inverse solver (given the signal, determine the source). Results: For a simple, small scale, and axisymmetric cylindrical gel configuration, and for a source at 500 Hz, the forward solve generates signals that agree with experimental data (using Kelvin-Voigt viscoelasticity). Also, with surface signals generated by simulated sources in this virtual environment the inverse algorithm is able to identify this source given only chest surface measurements, and an adequate initial datum from which to start the computation. Conclusions: While enormous challenges remain we have shown that this approach offers considerable promise in delivering a noninvasive diagnostic or screening tool.