Abstract
Radionuclide images by a scintillation camera are blurred mainly due to a low resolution of a detector itself and random incidence of γ-rays upon the camera. So it seems that the correction of radionuclide images might play an important role in the improvement of diagnosis based on radionuclide images. Scintillation camera data digitized by the data processing system CDS-4096 which was on-line connected with our scintillation camera (Pho-Gamma III) were off-line processed using FACOM-230/35 and NEAC-2230 computers installed at the data processing center in Kanazawa University.
Processed results were printed out by symbols of a lineprinter or punched out on paper tape. Moreover, the result on paper tape was fed back to CDS-4096 through a tapereader and was displayed on the cathode ray tube. In printing out of a lineprinter, the image size was transformed to the real size and the image contrast was increased by superposing multiple symbols of a typewriter.
And the distortion of image induced by non-uniformity of sensitivity of a detector was corrected by the data of the distribution of sensitivity checked in every time of an organ scintigraphy, using the scanning unit of241Am band source. Also, we always carried out smoothing in order to reduce a statistical fluctuation before enhancing the organ contour and defects in the organ.
In this smoothing, the weight coefficient matrix consisting of simple integers was applied. Then the following programs were studied. The organ contour was obtained by calculating differences of counts among each point and its surrounding points.
Multiple chosen level plotting of threshold was used in order to recognize defects. Several levels were chosen from 100 threshold levels of 1% degree and were plotted with superposing. In the image processed by such a method, defects could be detected quite apparently and their relative position in the organ also was known easily.
