Left ventricular segmentation from MRI datasets with edge modelling conditional random fields

Abstract

BackgroundThis paper considers automatic segmentation of the left cardiac ventricle in short axis magnetic resonance images. Various aspects, such as the presence of papillary muscles near the endocardium border, makes simple threshold based segmentation difficult.MethodsThe endo- and epicardium are modelled as two series of radii which are inter-related using features describing shape and motion. Image features are derived from edge information from human annotated images. The features are combined within a discriminatively trained Conditional Random Field (CRF). Loopy belief propagation is used to infer segmentations when an unsegmented video sequence is given. Powell’s method is applied to find CRF parameters by minimizing the difference between ground truth annotations and the inferred contours. We also describe how the endocardium centre points are calculated from a single human-provided centre point in the first frame, through minimization of frame alignment error.ResultsWe present and analyse the results of segmentation. The algorithm exhibits robustness against inclusion of the papillary muscles by integrating shape and motion information. Possible future improvements are identified.ConclusionsThe presented model integrates shape and motion information to segment the inner and outer contours in the presence of papillary muscles. On the Sunnybrook dataset we find an average Dice metric of 0.91±0.02 and 0.93±0.02 for the inner and outer segmentations, respectively. Particularly problematic are patients with hypertrophy where the blood pool disappears from view at end-systole.

Feature competition and partial sparse shape modeling for cardiac image sequences segmentation

“…The method [25] presented the edge modeling conditional random fields for the LV segmentation. The corresponding P2C errors of endocardium and epicardium were 1.57 mm and 1.78 mm respectively.…”

“…Table 1 provides some published results using the dataset [14]. The methods [14,23] segmented all the slices, the method [24] and our method segmented the mid slice, while the method [25] segmented the 6th slice. The methods [14,23,25] and our method segmented all the frames, while the method [24] only segmented frames in the end diastole (ED) and end systole (ES).…”

“…The methods [14,23] segmented all the slices, the method [24] and our method segmented the mid slice, while the method [25] segmented the 6th slice. The methods [14,23,25] and our method segmented all the frames, while the method [24] only segmented frames in the end diastole (ED) and end systole (ES). The methods [14,25] and our method segmented both the endocardium (endo) and epicardium (epi), while the methods [23,24] segmented only the endocardium.…”

“…The methods [14,23,25] and our method segmented all the frames, while the method [24] only segmented frames in the end diastole (ED) and end systole (ES). The methods [14,25] and our method segmented both the endocardium (endo) and epicardium (epi), while the methods [23,24] segmented only the endocardium. The method [14] used the 3D AAM segmentation as the initialization of the 2D ASM method to segment the LV.…”

Estimation of the Volume of the Left Ventricle From MRI Images Using Deep Neural Networks

“…It provides not only the diagnosis label for each case but also LV contours labeled by experts for some cases. It is the first systematically labeled dataset, which facilitates the application of machine learning algorithms on this problem, including deep neural network [5,6] and random forest [7]. Traditional learning-free methods can also benefit from it [8,9].…”