Research on Equivalence of SVD and PCA in Medical Image Tilt Correction

Abstract

In the process of medical imaging, often because of some disturbance, the medical images frequently have some undesirable tilt, which has costly negative effect on the following image alignment and fusion. In order to solve the tilt problem, Singular Value Decomposition (SVD) and Principal Component Analysis (PCA) are studied and their relationship between them is discussed, and then the medical image correction tilt process is divided into five main stages. Among these stages, the key tasks focus on finding the centroid and obtaining the tilt angle of a medical image. We use SVD and PCA to compute the eigenvectors of the coordinates of a medical image respectively to get the tilt angle. The experimental results reveal that the methods mentioned above are effective for correcting the tilt medical images and also prove the equivalence of SVD and PCA in medical image tilt correction.