Compton scattering imaging can demonstrate the electron density of a given object which could be important for medical diagnosis. However, clinical application of the technique has not yet been developed, because a clear image can not be obtained for several reasons. The image is seriously contaminated by multiple scattering which is inevitable with Compton scattering. The electron density in a small region of the object is not obtained correctly unless multiple scattering is handled carefully. In the present paper, methods to eliminate multiple scattering in Compton scattering imaging system are discussed. A general purpose Monte Carlo code called EGS4 was used to simulate the processes of an incident photon interacting with a substance inside an object. Simulations were performed using a water sphere phantom with a radius of 10 cm irradiated by a beam of monochromatic X rays at 100 keV. Scattered photons emitted from the phantom were observed and the ratio of the amount of single Compton scattering from a focal point to that of multiple scattering was calculated. To improve the image quality, elimination of multipe scattering in Compton scattering is indispensable. To this end,0 and 0 collimators and an energy window were proposed and their effects were estimated quantitatively by simulations. It was shown that by introducing of the proposed methods in the imaging system, the ratio of the Compton scattering signal to the multiple scattering noise, S/N, could be improved from 0.022 to 40. Quantitative results derived from the simulations may be useful for the design of a practical Compton scattering imaging system.
Nonuniform attenuation correction in SPECT was attempted to improve quantitative SPECT evaluation. Attenuation-coefficient maps required for the correction were obtained from X-ray CT image data. Two methods were used for the correction. One was Singh's method which was proposed for nonuniform attenuation correction. The other was Chang's method which was modified for nonuniform attenuation correction in the present work. The properties of the Singh method were studied by co mputer experiments for nonuniform mathematical phantoms and were compared with those of the modified Chang method. The two methods produced similar correction results. Problems in clinical applications o f attenuation correction using XCT data were studied. Three major problems were found: the determination of average energy of X-rays used to take XCT images, the effect of scattered photons, and the registration of anatomical positions between XCT image and the corresponding uncorrected SPECT image. Further study will be necessary to solve the problems.
We developed a ridge filter that enables to adjust the dose distribution in a spread out Bragg peak of proton beam by remote control. This device has 4 blocks a block of which is a quarter of a conical shape of MiX-DP block. Each pair of two blocks is put on an each line perpendicularly intersecting each other. The intersecting point of these two lines is coincided with an axis of proton beam. We could flatten the dose distribution of spread out Bragg peak of proton beam by adjusting the distance between two blocks of each pair. We, moreover, can adjust it if necessary, for example, in a case that relative biological effectiveness in spread out Bragg peak increases with depth.
Stopping power and range functions have been calculated for carbon ions and are presented in the form of tables. With a suitable interpolation scheme, intermediate values can be obtained. The uncertainty of the values is of the order of 2% at high ion speeds, but much larger at small speeds.