“…The other is necessary to use some training datasets, and is called segmentation with strong prior [8]. The examples of segmentation algorithms are thresholding [8,9], region growing [10], dynamic programming (DP) [11], deformable models [12], graph cuts [12], active contour models (ACM) [13][14], level-set [15], KNN classifier [16], convex relaxed distribution matching [17], robust adaptive Gaussian regularizing Chan-Vese (CV) model [18], and clustering [19]. Most of the previous works were focused on automatic LV segmentation for evaluating the cardiac function, but only a few works were focused on cardiac T2* estimation.…”