Cardiac Function Evaluation

Magnetic resonance imaging (MRI) has become a key medical image modality. It can offer high and selective contrast and high spatial resolution without using ionizing radiations. Besides, dynamic MRI allows the physician to examine time dependent phenomena, such as cardiac movement or contrast concentration in tissues.

In our lab, we have developed several algorithms and techniques that allows us to extract valuable information about the cardiac function, such as myocardium strain or perfusion concentration.

• Local parameters derived from strain tensor

There is a wide spectra of cardiopathies in which myocadial function is altered, the local evaluation of which allows to determine wich regions of the myocardium are compromised. This way, it is now possible to emit a more precise diagnosis that allows a more effective therapy.

In order to estimate the myocardial function locally, an automatic method based on the myocardial strain tensor computation from cardiac magnetic resonance acquisitions has been developed. Tensors are computed by combining the properties of a physical model of the myocardial muscle deformation and the information provided by the myocardium acquired images.

So far, this method has been applied to the characterization of the myocardial deformation parameters in diverse pathologies.

• Perfusion Quantification and Myocardial Viability

Perfusion pharmacokinetic analysis using cardiac magnetic imaging allows early detection and diagnosis of regional defects in myocardial irrigation which can subsequently lead to fibrosis or scarring of tissue. Delayed enhanced cardiac MRI can be used in this scenario —as well as for various types of cardiopathies— to know the transmural extent damaged tissue, resulting of great interest for therapeutic decision making.

We have developed a collection of methods to segment the myocardium, correct the misalignment of the sequences due to patient breathing and to robustly estimate the pharmacokinetic model parameters. In order to quantify the transmurality of the lesson, an initial segmentation of the scarred tissue is obtained. Our developed methods calculate simultaneously the thickness both of the scar and the healthy myocardium. Dense 3D maps of local transmurality are obtained from the calculated values.

It provides a comprehensive tool for pharmacokinetic analysis of myocardial perfusion and the transmural scar tissue from magnetic resonance, for which there are no solutions on the market nowadays.