Detection of two-dimensional respiratory movement and its correction on the scintillation camera images

Processing of images obtained with a scintillation camera by the on-line computer was described for correction of image distortion due to the respiratory movement. Serial image data sampling and calculation of the center of gravity for each frame allowed the detection of two-dimensional respiratory movement and correction of respiratory movement by superimposition of the frames according to each center of gravity. Experiments with liver phantom and respiration simulator showed that the optimum sampling interval was one-fourth of the respiration cycle in conjunction with siatistical fluctuation in calculation of the center of gravity and precise decision of magnitude of respiratory movement. Results of the calculation well agreed with the actual movement of the phantom. The superimposed images in phantom and patients also well agreed with the static images without respiratory movement. The serial calculations of center of gravity were applied to the verticaliy divided sub-images and possibility of correction was suggested in such a case that the magriiude of movement was not even throughout the organ. In the liver image, the horizontal movement falled to be differentiated from the statistical fluctuation. The correction was not only useful for increased detectability of small lesion but also essential for computer image restoration asmore » a pre-processing. (auth)« less