Simulation and in-vivo validation methods for the assessment of global and regional myocardial function from feature tracking and cardiac tagging

Project developed by final year PhD QMUL student Haifa Almutairi under the supervision of Prof. Steffen Petersen, Dr. Redha Boubertakh and Dr. Marc Miquel.

The British Heart Foundation states that CVD causes more than 25% of all deaths in the UK. Cardiac-MR is considered the gold standard for quantification of cardiac function using both global volumetric measurements (e.g. ventricular and atrial volumes, ejection fraction…) and global and regional motion measurements (e.g. strain, strain rate…).

Cardiac-MR feature tracking (CMR-FT) techniques are commonly used methods to extract motion information from cardiac cine images.

Strain Curves
Figure 1. Strain curves from TOMTEC software in a (a) short-axis view and (b) vertical long axis 2 channel view.

However, there is a lack of validation of feature tracking techniques. Most studies were carried out to compare feature tracking to other imaging modalities (e.g. echocardiography) or MRI techniques (CMR tagging) where the true cardiac motion is unknown. There are at least 3 commercial use software packages with no corresponding published phantom validation studies. The main aim in this project was to validate cardiac-MR feature tracking methodologies by comparing several commercial software packages using healthy and diseased populations. Cardiac-MR feature tracking was also compared to cardiac-MR tagging, the gold standard for measuring cardiac motion. Lastly, a gold standard numerical phantom (i.e. known motion) was developed to assess sources of differences between commercially available software packages.

Developed methodologies could be extended to other MR applications and imaged body parts.