Physiome Model Based Cardiac Information Recovery Under Multiframe State-Space Framework

In patient-specific cardiac information recovery, meaningful a priori models for constraining the sparse and noise-corrupted measurements, and also the optimal criteria for coupling them together, are essential for obtaining more reliable estimates. Although the extensively used biomechanical models give promising results, it cannot account for the active components of the myocytes. In view of this, we propose to adopt a cardiac physiome model, which is composed by a cardiac electric wave propagation model, an electro-mechanical coupling model, and a biomechanical model, for a more complete modeling of the cardiac physiology. Under the multiframe state-space filtering framework, the physiome model is coupled with the measurements to provide the optimal estimates of the cardiac information. Experiments have been performed on both synthetic and MRI data to illustrate its abilities and benefits