Quantitative Accuracy of Low-Count SPECT Imaging in Phantom and In Vivo Mouse Studies

We investigated the accuracy of a single photon emission computed tomography (SPECT) system in quantifying a wide range of radioactivity concentrations using different scan times in both phantom and animal models. A phantom containing various amounts of In-111 or Tc-99m was imaged until the activity had decayed close to background levels. Scans were acquired for different durations, employing different collimator pinhole sizes. VOI analysis was performed to quantify uptake in the images and the values compared to the true activity. The phantom results were then validated in tumour-bearing mice. The use of an appropriate calibration phantom and disabling of a background subtraction feature meant that absolute errors were within 12% of the true activity. Furthermore, a comparison of in vivo imaging and biodistribution studies in mice showed a correlation of 0.99 for activities over the 200 kBq to 5 MBq range. We conclude that the quantitative information provided by the NanoSPECT camera is accurate and allows replacement of dissection studies for assessment of radiotracer biodistribution in mouse models.