The lung cancer is very difficult to treat when condition of disease reaches an advanced stage. Therefore, the early detection and the early treatment by the mass cscreening are important. Now, the3 mass screening using the chest X-rays film is performed, and its detection rate is low. Recently, mass screening for lung cancer started using helical CT. However, since each subject has about 30 images, there is concern about the increase ofa burden to a physician. This comparative reading system solves difficulties of efficient display with the past and present images. But, automatic slice-image-matching is difficult by computer, since the states of the lungs at the time of photography differ from each other. This research analyzed change of the lungs between images with time and proposed automatic slice image mtching algorithm for comparative reading.