Quantized Segmentation of Fibrotic Tissue of Left Atrial from Delay-Enhancement MRI Images Using Level-Set and Graph-Cut

The pathogenesis of atrial fibrillation (AF) is closely related to the fibrotic tissues in left atrial (LA). Delay-enhancement magnetic resonance imaging (DE-MRI) has been widely used in the ablation of atrial fibrillation, which can accurately describe the distribution of myocardial fibrosis and postoperative scars. Combining EM algorithm, level-set and graph-cut, this paper proposes a method to segment the fibrotic tissues and postoperative scars and also quantify their proportion in left atrial from DE-MRI. In 4 clinical cases, our method can accomplish the extraction of heart, the segmentation of left atrium and sequentially the quantification of the fibrotic tissues nearby with little manual intervention. Experimental results show that accurate segmentation of LA is achieved in 55 slices with 96 slices containing LA among 4 cases in total. With manual correction in the rest slices, the final results about the proportion of fibrotic tissues in LA are 14.78%, 21.02%, 25.17%, 14.77% respectively which are consistent with the clinical diagnosis. Evaluated by the clinician, our method is robust against different resolution and can provide auxiliary function for ablation of AF.