Semiautomatic detection of myocardial contours in order to investigate normal values of the left ventricular trabeculated mass using MRI

Abstract

We propose a semiautomatic method based on region growing, active contours, and thresholding to calculate the NC:C mass ratio. This method is highly reproducible and might help in the diagnosis of LV noncompaction cardiomyopathy. J. Magn. Reson. Imaging 2016;43:1398-1406.

Fully automated segmentation of the left ventricle in cine cardiac MRI using neural network regression

“…The LVSC and Kaggle databases were sourced from a variety of imaging centers and scanner types, leading to a heterogeneous mix of imaging protocols and parameters. For the LVSC database, the range (mode) of imaging parameters were: echo time (TE) 0.96–2.98 (1.13) msec, repetition time (TR) 2.50–79.1 (59.2) msec, flip angle 25–90° (45°), in‐plane (x/y) resolution 0.68–2.14 (1.56) mm, slice thickness 6–10 (8) mm, number of slices 5–17, number of cardiac phases 18–35 . For the Kaggle database, the range (mode) of imaging parameters were: TE 1.04–1.54 (1.19) msec, TR 14–54.7 (34.2) msec, flip angle 35–79° (50°), in‐plane (x/y) resolution 0.59–1.95 (1.41) mm, slice thickness 5–11 (8) mm, number of slices 2–21 (10), number of cardiac phases 25–30 .…”

“…Techniques used include general image processing‐based methods such as intensity distribution modeling of the LV tissue and blood pool; deformable models such as active contours targeting the myocardium boundaries; statistical shape and appearance models; and anatomical atlas‐based registration . The attributes of published algorithms vary, ranging from semiautomated to fully automated; endocardium only to complete myocardium; mid‐slices only to full coverage from apex to base (mitral valve); and dual ED/ES phases only to all cardiac phases . Fully automated algorithms are inherently superior in terms of convenience, as well as their elimination of subjective interobserver variability.…”

Influence of observer-dependency on left ventricular hypertrabeculation mass measurement and its relationship with left ventricular volume and ejection fraction – comparison between manual and semiautomatic CMR image analysis methods

“…In turn, the major advantage of the presented computed method is its semi-automatic character, exclusion of the blood pool from analysis and simplification related to the unanimous inclusion of PMs into the trabecular mass. Bricq et al previously introduced the semiautomatic assessment of the trabeculated and compacted LV mass, but the authors excluded both PMs and ITB from NCM [20,21]. It is essential to consider the precision and accuracy of the different methods of NCM analysis.…”

The Assessment of left ventricular Function in MRI using the detection of myocardial borders and optical flow approaches: A Review

“…In clinical practice, some radiologists use only the myocardial contours of the end diastole image (corresponding to the telediastolic moment when the left ventricular volume reaches its maximum) and the end systole image (where the volume of the LV is at a minimum) to compute LVEF. To overcome these limitations, several researchers have developed semi-automatic or automatic approaches for the detection of myocardial contours [4,15]. We distinguish the approaches based on the intensity of the pixels which were initially proposed for the segmentation of the endocardial contours.…”