Correction of liver displacement due to irregular respiration for SPECT images obtained using a multiple short-time acquisition with breath-holding technique

Recently, a new method for acquiring single photon emission computed tomography (SPECT) data during breath-holding with a high-speed rotation (HSR-SPECT) technique was applied to hepatic-function scintigraphy. This technique can suppress motion artifact caused by respiration. However, it is challenging to apply this technique to patients with respiratory problems. Our purpose in this study was to analyze projection data and correct liver displacement due to irregular respiration during technetium-99m-diethylenetriaminepentaacetic acid galactosyl human serum albumin (99mTc-GSA) liver SPECT scans. We calculated cross-correlation functions (CCFs) for both intra- and inter-projections to correct irregular projections by shifting them in the longitudinal direction enough to maximize the CCFs. We integrated all of the shifted projections and created a synthesis projection which was used to reconstruct SPECT images. To confirm the effectiveness of our method, we did both a phantom study and a clinical retrospective review. In the phantom study, blurring artifacts due to displacement were reduced using our method. Post-processing using our method showed improvements in the normalized mean square error values ranging from 0.18 to 0.06. Furthermore, our method showed no disadvantage or miss-correction if we applied our method to data without displacement. In the clinical review, our method improved the image clarity for patients who could not repeat consistent respiration. In both the phantom simulation and the clinical application, the results showed the usefulness of our method. The efficiency of our method has the potential to reduce the displacement problem caused by respiration and provides excellent matching SPECT/CT fusion images.