Abstract
Image contrast defined by the count density of nuclear medicine image is often used to evaluate abnormal accumulation of radioisotope (RI) due to disease. However, the image is not always a true radioactivity distribution of the patient, because of the presence of many factors that degrade the image quality. These are inadequate resolution of the scintillation camera, and absorption and scatter of γ-rays. Therefore, it is important to examine how well the image reflects the object radioactivity distribution. In the present paper, the relation between object contrast, defind by RI distribution, and image contrast, difined by count density, was investigated experimentally. The phantom 1 consists of a cube with low background activity and a sphere or circular disk with high activity. Phantom 2 (cold case) consists of a cube with high activity and a sphere without RI activity. These phamtoms were imaged with a scintillation camera and the image contrast (count density of the sphere or disk/background count density) was calculated. Correlation between the object contrast and image contrast was linear, supporting its use to evalute the abnormal accumulation of RI in a patient. However, on account of the effect of scatter, the absolute value of image contrast was somewhat smaller than the object contrast. It should then be important, considering above results, to use the image contrast in clinical study.
