Automatic segmentation, detection and quantification of coronary artery stenoses on CTA

Abstract

Accurate detection and quantification of coronary artery stenoses is an essential requirement for treatment planning of patients with suspected coronary artery disease. We present a method to automatically detect and quantify coronary artery stenoses in computed tomography coronary angiography. First, centerlines are extracted using a two-point minimum cost path approach and a subsequent refinement step. The resulting centerlines are used as an initialization for lumen segmentation, performed using graph cuts. Then, the expected diameter of the healthy lumen is estimated by applying robust kernel regression to the coronary artery lumen diameter profile. Finally, stenoses are detected and quantified by computing the difference between estimated and expected diameter profiles. We evaluated our method using the data provided in the Coronary Artery Stenoses Detection and Quantification Evaluation Framework. Using 30 testing datasets, the method achieved a detection sensitivity of 29% and a positive predictive value (PPV) of 24% as compared to quantitative coronary angiography (QCA), and a sensitivity of 21% and a PPV of 23% as compared manual assessment based on consensus reading of CTA by 3 observers. The stenoses degree was estimated with an absolute average difference of 31%, a root mean square difference of 39.3% when compared to QCA, and a weighted kappa value of 0.29 when compared to CTA. A Dice of 68 and 65% was reported for lumen segmentation of healthy and diseased vessel segments respectively. According to the ranking of the evaluation framework, our method finished fourth for stenosis detection, second for stenosis quantification and second for lumen segmentation.

Precise Lumen Segmentation in Coronary Computed Tomography Angiography

“…in the presence of calcified, non-calcified and mixed plaques. This can be partially accounted for by explicitly modeling or suppressing calcified plaque before segmentation [7,9,10]. Many of the proposed approaches also involve postprocessing and refinement of the segmentation results for fixing artifacts.…”

“…Many of the proposed approaches also involve postprocessing and refinement of the segmentation results for fixing artifacts. Shazad et al [10], for example, propose a voxel-based graph-cut segmentation followed by a radial resampling and smoothing which patches non-tubular segmentation results (e.g. dissected lumen segmentations) and artifacts due to the voxel-level accuracy.…”

Automatic detection of aorto-femoral vessel trajectory from whole-body computed tomography angiography data sets

“…For our purposes, we distinguished two background categories: high intensity tissue (calcifications, bones, and metal artifacts), and low intensity tissue (muscles, fat, liver etc.). We masked out the surrounding background by an adaptive threshold method similar to Shahzad et al [15].…”

Segmentation and Visualization of Human Coronary Artery Trees from CTA Datasets

“…Several algorithms have been developed to segment and visualize vessels in three dimensions [3], for instance, level set method [10], active contour algorithm [14], vesselness measurement [11], [12], expectation maximization estimation algorithm [15], moment-based shape analysis for voxel clusters [16], shape model based algorithms such as tubular model in three dimensions [17], the algorithm of combining graph-cuts and robust kernel regression to segment coronary lumens [18], [19]. The standard marching cube algorithm [21] generates a high resolution isosurface with an isovalue of image intensity to represent the object's surface.…”

“…The computed mean value plus three standard deviations of the intensities is the threshold value to recognize the calcified plaques [25]. Stenoses or soft plaques are studied to be detected with profiles of artery lumen sectional area or vessel radiuses along the vessels' centerlines [15], [27], [30]. The detections of the artery lumen, calcified and soft plaques or stenoses are clinical useful [31][32][33].…”