Julia Lockwood

Dynamic SPECT Measurement of Absolute Myocardial Blood Flow in a Porcine Model

Absolute myocardial blood flow (MBF) and myocardial flow reserve (MFR) provide incremental diagnostic and prognostic information over relative perfusion alone. Recent development of dedicated cardiac SPECT cameras with better sensitivity and temporal resolution make dynamic SPECT imaging more practical. In this study, we evaluate the measurement of MBF using a multipinhole dedicated cardiac SPECT camera in a pig model of rest and transient occlusion at stress using 3 common tracers: 201Tl, 99mTc-tetrofosmin, and 99mTc-sestamibi. Methods: Animals (n = 19) were injected at rest/stress with 99mTc radiotracers (370/1,100 MBq) or 201Tl (37/110 MBq) with a 1-h delay between rest and dipyridamole stress. With each tracer, microspheres were injected simultaneously as the gold standard measurement for MBF. Dynamic images were obtained for 11 min starting with each injection. Residual resting activity was subtracted from stress data and images reconstructed with CT-based attenuation correction and energy window–based scatter correction. Dynamic images were processed with kinetic analysis software using a 1-tissue-compartment model to obtain the uptake rate constant K1 as a function of microsphere MBF. Results: Measured extraction fractions agree with those obtained previously using ex vivo techniques. Converting K1 back to MBF using the measured extraction fractions produced accurate values and good correlations with microsphere MBF: r = 0.75–0.90 (P < 0.01 for all). The correlation in the MFR was between r = 0.57 and 0.94 (P < 0.01). Conclusion: Noninvasive measurement of absolute MBF with stationary dedicated cardiac SPECT is feasible using common perfusion tracers.

Sci—Fri AM: Imaging — 09: Serial estimation of cross‐talk for correction in dual‐isotope imaging with dynamic tracers

The recent radioisotope shortage has led to interest in non-Tc99m-based tracers. We have developed a novel I-123-labelled myocardial perfusion imaging tracer. We compare the I123-tracer to the clinical standard of Tc99m tetrofosmin in vivo in a rat model using a small-animal SPECT/CT camera. SPECT distinguishes different isotopes based on the different energies of the emitted gamma rays and thus allows simultaneous comparison of two tracer distributions in the same animal. Dual-isotope imaging is complicated by cross-talk between the energy windows of the isotopes. Standard energy-window-based correction methods are difficult to employ because of the proximity in energy of Tc99m (140keV) and I123 (159keV). Imaging the second tracers energy window prior to its injection provides an estimate of the cross-talk. However, this estimate is only accurate if the tracer distribution is static. We use serial imaging prior to the introduction of the second tracer to estimate the dynamics of the first tracer and interpolate the cross-talk images to provide a more accurate correction. We used rat models of myocardial disease (n=3). I123 tracer was injected and imaged for one hour at 20min intervals. The Tc99m tetrofosmin was then injected and 30min later, a dual-isotope image was obtained. The impact of this approach is assessed by comparing the differences in the Tc99m-tetrofosmin image using this method with correction by simple correction for physical decay. The interpolative approach improves the accuracy of the correction by 2%-5% and thereby enhances the comparison of the two tracers.

Poster — Thur Eve — 53: Reproducibility of TI-201 for Cardiac Micro SPECT Imaging with a Rat Model

Serial imaging with micro SPECT/CT is an important 3D in vivo technique used to study heart disease in small-animal models and develop new therapies and new radiotracers. Measurement of myocardial perfusion uniformity (PH), ejection fraction (EF), end systolic volume (ESV) and, end diastolic volume (EDV) in rat model are important indices of heart function that can be derived from SPECT images. Knowledge of the uncertainty of these measurements is critical to discerning true changes and to determining sample sizes. The aim of this study is to produces the inter- and intra- subject reproducibility of left ventricular volumes, ejection fraction and perfusion homogeneity with Tl-201, a common cardiac perfusion tracer, in a rat model using micro SPECT/CT. The scanner has four heads with nine pinholes each (total 36 pinholes). Three healthy rats were injected with 0.5mCi of Tl-201 and scanned weekly for five weeks. Each scan lasted 30 min and began 30 min post-injection. The images were reconstructed and processed with a clinical software package (4DM-SPECT) to get the heart functions (EDV, ESV, EF and PH). EDV and ESV were corrected for changes in rat weight. The standard deviation in the measured values (across scans) was 5.88% for EDV and 6.90% for ESV, 3.2% for EF, and 6% for PH. The reproducibility between rats was similar.

Poster - Thur Eve - 69: Attenuation Correction for 99m Tc-Tetrofosmin Micro SPECT/CT Cardiac Measurements in Rats

Objective: Small animal microSPECT is an important pre‐clinical imaging modality. However, the quantitative accuracy may be somewhat limited by photon loss due to attenuation in the body of the subject. Our goal is investigate the effects of including attenuation correction in the reconstruction of rat SPECT/CT data, and to assess the impact on the quantitative uptake in the myocardium and the uniformity of the distribution of uptake. Method: A rat was injected with ∼2.0 mCi of 99mTc‐tetrofosmin (Myoview) and scanned with a nanoSPECT scanner (BioScan, Washington, DC). Scan parameters were as follows: single, 1.5 mm pinhole, 4 hour scan duration commencing 30 minutes post injection, 48 projections, and circular scan. The rat was also scanned with the built‐in CTscanner (45 keV, cone beam). The CT data were reconstructed using filtered backprojection and rescaled to generate a linear attenuation map for the 140 keV 99mTc photons. The SPECT data were reconstructed using an iterative ordered subsets expectation maximization (OSEM) algorithm. Iterative reconstruction allows photon attenuation to be modelled as part of the reconstruction algorithm, and the reconstruction was done both with and without including attenuation. Results: As expected, inclusion of attenuation correction in the reconstruction increased by about 30% the total observed activity in the heart. However, we found no evidence that ignoring the attenuation correction introduces any significant artefacts in the uniformity of the reconstructed activity map. Conclusions: Including attenuation correction in the reconstruction of 99mTc cardiac images has little impact on the measured distribution of activity in the myocardium.