Cardiac image segmentation by random walks with dynamic shape constraint

Abstract

The quantitative analysis of the left ventricle (LV) contractile function is one of the key steps in the assessment of cardiovascular disease. Such analysis greatly depends on the accurate delineation of LV boundary from cardiac sequences. However, segmentation of the LV still remains a challenging problem due to its subtle boundary, occlusion, and image inhomogeneity. To overcome such difficulties, the authors propose a novel segmentation method by incorporating a dynamic shape constraint into the weighting function of the random walks segmentation algorithm. This approach involves iterative updates on the intermediate result to achieve the desired solution. The inclusion of a shape constraint restricts the solution space of the segmentation result to handle misleading information that may come from noise, weak boundaries and clutter, leading to increased robustness of the algorithm. The authors describe the details of the proposed method and demonstrate its effectiveness in segmenting the LV from real cardiac magnetic resonance (CMR) image sets. The experimental results demonstrate that the proposed method obtains better segmentation performance than the standard method.

Mean curvature and texture constrained composite weighted random walk algorithm for optic disc segmentation towards glaucoma screening

“…The DIARETDB1 dataset consists of 89 colour where G R and D R correspond to the groundtruth and segmented OD regions, respectively. The performance of the CTCRW algorithm for OD segmentation is first compared with the classical RW [29] and its recent variations applied on medical images [30,31]. In order to make an unbiased performance comparison, the preprocessing and seed initialisation process are kept identical for each method.…”

“…An extra penalty term is utilised in the weighting function of [31] which penalises the dispersion of Gaussian-filtered intensities. The segmentation algorithms in [30,31] are referred as RWDT (RW with distance transform) and RWP (RW with penalty function) here.…”

“…In addition, it avoids the trapping at local minima in deformable models and small cut problem in graph cuts. RW-based segmentation was applied successfully in biomedical images such as left ventricle segmentation in cardiac magnetic resonance images [30] and tumour segmentation in brain and liver images [31].…”

Left Ventricular Wall Stress Is Sensitive Marker of Hypertrophic Cardiomyopathy With Preserved Ejection Fraction

“…First, our approach to 3D LV reconstruction relies on manual delineation of endocardium and epicardium contours obtained from CMR images. This is time consuming and may be replaced by automatic segmentation techniques (Petitjean and Dacher, 2011 ; Kang et al, 2012 ; Yang et al, 2016 ).…”

Deep convolutional neural networks for automatic segmentation of left ventricle cavity from cardiac magnetic resonance images

“…U-Net-L is using localisation-segmentation scheme, while U-Net is without this scheme Table 2 Comparison of LV endocardium segmentation performance in terms of DM between the proposed U-Net-DL and the state-of-art methods on the datasets from MICCAI 2009 LV Challenge website [34]. The results of the state-of-art methods are taken from [28,38] Run time per frame, s DM (mean ± std) APD (mean ± std), mm Good contours*, % volumetric medical image analysis [31,46,47]. Though 3D neural networks have the potential to solve the difficulties occurred in cardiac segmentation, the practical application is limited by the large gap (slice thickness) of CMR imaging as well as the lack of enough amount of patient subjects.…”