ALVLS: Adaptive local variances-Based levelset framework for medical images segmentation

Xiu Shu, Yunyun Yang, Jun Liu, Xiaojun Chang, Boying Wu

Abstract: Highlights•An adaptive local variances-based coefficient is proposed to distinguish noise points and the edge of an object, which improves the segmentation accuracy of medical images with intensity inhomogeneity and noises.•The edge detection function of the LRFLSM model is reverted to a simple detection function so that the edge detection function does not need to update with iteration.•The two-layer local variances-based level set framework for segmenting left ventricles and left epicardium simultaneously.•Experimental results for medical images and synthetic images show the desirable performance of the ALVLS model in accuracy, efficiency, and robustness to noise.