Methods: Since dynamic SPECT computes the MBF from TACs, it was envisioned that the phantom must mimic physiological TACs. In this case, radionuclide temporal behavior in both the left ventricle cavity (LV) and in the myocardium regions of the phantom should allow rapid increase and decrease of radiotracer concentration in the left ventricle cavity and monotonic increase of radiotracer concentration in the myocardium region. To achieve these goals, the newly designed phantom consists of the following components: (1) Two nested flexible silicon membranes – the internal volume representing the LV and the intermediate volume between both membranes representing the myocardium region surrounding the LV. (2) The dimensions of the membranes resemble the endocardium and epicardium of an actual human heart. (3) Multiple ports for both regions for injection and flushing of radiotracer and saline/water. (4) A large diameter port representing the aorta connected to the base of the LV region, allowing for significant stroke volume at each "cardiac beating" using a pulsatile pump.
Results: Temporal variation of radiotracer concentration in both of the phantom's volumes was obtained by electronically-controlled syringe injectors, typically injecting 0.1-0.5 mCi of 99mTc-SESTAMIBI directly into the phantom. The injection profiles were designed to resemble the TACs observed in human patients: A high-concentration activity bolus arriving into the LV and rapidly dropping after leaving the heart through the aorta, followed by monotonic increase of the tracer concentration until it reaches its final state. A comparison between dynamic SPECT measurements of the phantom and a human patient is shown in the attached figure.
Conclusions: A newly developed cardiac phantom for dynamic SPECT/PET measurements was presented. The phantom is able generate reproducible, reliable results to investigate key parameters of dynamic SPECT/PET systems, namely, the accuracy and reproducibility of TACs and their propagated effect on the computed MBF.