Tissue Cell Boundaries Detection based on Curvelet-based Snake Model in Electrorotation Bio-chip Control System

Abstract

During cell motility tracking process, to suppress background interference and separate the clustering cells contour, a novel cell boundary feature extraction algorithm based on snake model and curvelet transform(CT) is proposed. The CT has higher time-frequency resolution, high degree of directionality and anisotropy. Microscope cell image is denoised by discrete CT to improve SNR. Then curvelet-based Snake model in multiscale space is applied to identify cell contour, which obtains cell boundaries under the effect of both the original image force and the modified nonlinear distance image force. It is effective for electrorotation cell contour extraction corrupted by noise, with weak edges. Experimental results show good performances of the proposed method to extract the shape of edges and quantification of cell motility.