|Ph.D Student||Tsadok Yossi|
|Subject||Cardiac Anatomy/Substrate Analysis Using Cardiac-MRI|
Images and Echocardiographic Data
|Department||Department of Biomedical Engineering||Supervisor||Professor Emeritus Dan Adam|
Cardiac magnetic resonance (CMR) imaging has emerged as an established technique providing accurate and reliable means of assessing function and anatomy of the heart. Analysis of CMR datasets requires segmentation of the myocardial boundaries in cine CMR and in Late Gadolinium Enhancement (LGE)-CMR.
An automatic algorithm is proposed for segmentation of both endocardial and epicardial boundaries. In addition, a novel postprocessing method was developed for assessment of longitudinal strain in standard cine CMR datasets. The segmentation algorithm is based on edge detection in initial frame, following by applying a non-rigid registration process between subsequent frames for tracking the segmentation from the initial frame throughout the entire sequence. The later algorithm was used also for evaluation of midmyocardial longitudinal strain in long-axis view.
Validation of the segmentation of the CMR data was performed using manual tracing of the boundaries, which shows excellent correlation for measuring clinical parameters: r=0.985 (p<0.001) and r=0.982 (p<0.001) for end-diastolic volume and end-systolic volume, respectively.
The algorithm for assessment of longitudinal strain was validated with tagged -Magnetic Resonance Imaging and Speckle Tracking Echocardiography. A comparison of peak systolic strain measurements yielded a strong correlation with these methods. In addition, a comparison with scar extent, as measured by LGE-MRI, shows discriminative power to identify non-transmural and transmural infarcted myocardium.
In conclusion, a tool for segmenting the myocardial boundaries in the long-axis views is proposed, which works well, as demonstrated by the validation performed using a clinical dataset. The study also showed a novel off-line postprocessing method for segmental longitudinal strain analysis on cine CMR data. This tool allows good discrimination between different transmurality states in patients with acute myocardial infarct.