Improved image quality in small-animal gated cardiac SPECT using non-rigid motion compensation

Abstract

101

Objectives: The objective of this work was to improve small-animal gated cardiac SPECT images using a new non-rigid motion estimation method combined with temporal summation.

Methods: Gated cardiac images of a mouse were obtained using a stationary SPECT system with 75 focusing 0.6mm pinholes (U-SPECT-II, MILabs, The Netherlands). The non-rigid motion between reconstructed image frames was estimated using a new method that generalizes a standard optical flow algorithm into a successive quadratic approximation framework. The method has demonstrated improved motion estimation accuracy. Using the estimated frame-to-frame motion, all frames were deformed to a reference frame and summed to achieve reduced image noise with limited motion blur.

Results: The proposed method generated images of superior image quality to the original, uncompensated images. Specifically, (i) the right ventricular wall now stayed visible throughout the cardiac cycle, (ii) a lower overall noise level was achieved, and (iii) the papillary muscles showed improved definition.

Conclusions: We have developed a method for improving the image quality of small-animal gated cardiac SPECT images using non-rigid motion estimation combined with temporal summing. Future work will concentrate on implementation of the method into a maximum a posteriori reconstruction framework.

Research Support: This work was supported by the National Inst. of Health under Grant RO1HL073336, and the United States Army Medical Research and Material Command under Award No. W81XWH-04-1-0594.