Abstract
Segmentation of the left atrium (LA) from cardiac magnetic resonance imaging (MRI) datasets is of great importance for image guided atrial fibrillation ablation, LA fibrosis quantification, and cardiac biophysical modelling. However, automated LA segmentation from cardiac MRI is challenging due to limited image resolution, considerable variability in anatomical structures across subjects, and dynamic motion of the heart. In this work, we propose a combined random forests (RFs) and active contour model (ACM) approach for fully automatic segmentation of the LA from cardiac volumetric MRI. Specifically, we employ the RFs within an autocontext scheme to effectively integrate contextual and appearance information from multisource images together for LA shape inferring. The inferred shape is then incorporated into a volume-scalable ACM for further improving the segmentation accuracy. We validated the proposed method on the cardiac volumetric MRI datasets from the STACOM 2013 and HVSMR 2016 databases and showed that it outperforms other latest automated LA segmentation methods. Validation metrics, average Dice coefficient (DC) and average surface-to-surface distance (S2S), were computed as 0.9227 ± 0.0598 and 1.14 ± 1.205 mm, versus those of 0.6222–0.878 and 1.34–8.72 mm, obtained by other methods, respectively.
Highlights
Atrial fibrillation (AF) is the most common cardiac electrical disorder and a major cause of stroke [1]
Accurate segmentation of the left atrium (LA) anatomy from MR images is of great importance for ablation guidance during the therapy procedure, automatically quantifying the LA fibrosis which is highly associated with postablation AF recurrence [3] and creating cardiac biophysical models [4, 5]
In order to better understand the influence of the different stages of the proposed method, Figure 5 shows the outcomes of the volume-scalable active contour model (ACM) without shape constraint, the concatenated random forests (CRFs) without contour refinement, and the integrated CRFs and volume-scalable ACM on case B003 of the STACOM database, respectively
Summary
Atrial fibrillation (AF) is the most common cardiac electrical disorder and a major cause of stroke [1]. Ablation of AF has become a commonly performed therapy procedure in many major hospitals throughout the world [2]. Accurate segmentation of the LA anatomy from MR images is of great importance for ablation guidance during the therapy procedure, automatically quantifying the LA fibrosis which is highly associated with postablation AF recurrence [3] and creating cardiac biophysical models [4, 5]. The LA areas can be defined through intensity gradients between the blood pool and surrounding tissues, the adjacent anatomical structures, such as other cardiac chambers, the descending aorta, and the coronary sinus, present signal intensities similar to that of the blood pool, and even manual segmentations by expert raters may show significant variations. The boundary between the LA and the left ventricle is difficult to define due to the different opening positions of the mitral valve (MV) leaflets [7]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
