Identifying optimal SPECT acquisition for liver imaging using 99mTc-MAA

Objectives To identify the optimal SPECT image acquisition time per view for liver imaging using 99mTc-MAA.

Methods Nominal 5 mCi of 99mTc-MAA was injected via the hepatic artery of the patient to assess perfusion of liver lesions as part of 90Y-microsphere therapy planning. Gated SPECT data was acquired and rebinned to generate 5 separate SPECT acquisition times: 34 s/view 28 s/view, 24 s/view, 20 s/view, and 16 s/view, respectively. Volumes of interest (VOI) were drawn over liver lesions and right kidney (background) for signal analysis.

Results Imaging acquisitions of 16 seconds and 20 seconds per view demonstrated little variation in signal to noise ratio (<2%), and the percent variation of liver counts per ml vs. time ratio was less than 2.52% . SPECT images acquired at 24 s/view and greater showed signal overflow in 2 of the 6 (33%) patients imaged. Using the tumor counts per mL vs. time ratio we were able to determine that count variation greater than 10% resulted in overflow of counts in the targeted region.

Conclusions Statistical data acquired identifies no significant change in signal to noise ratio or liver counts per mL to time ratio in imaging azimuths of 16 seconds and 20 seconds and potential SPECT signal overflow for times 24 s/view and greater . The results suggest SPECT acquisition time of 16 s/view is optimal for liver imaging using 99mTc-MAA without statistically sacrificing image quality