Abstract
The temporal subtraction image, which is obtained by subtraction a previous image from a current one, is an important technique for visual screening for detecting of the temporal changes such as new lesions. In this paper, we propose a new method for temporal subtraction image which is obtained two difference time series. We describe a method to reduce of computational times by use of an octree algorithm on the thoracic MDCT images. To achieve a subtraction image, we perform a global matching technique by use of affine transformation from a previous and current image sets in the initial step. In the next stage, new image registration technique for local matching is performed by use of octree algorithms on the volume of interest in order to reduce the computational times. Some experimental results are shown with discussions.
