Degradation of myocardial perfusion SPECT images caused by contaminants in thallous (201Tl) chloride

Purpose Thallous (Tl-201) chloride is a single-photon emission computed tomography (SPECT) tracer mainly used for assessing perfusion and viability of myocardial tissue. Tl-201 emits X-rays around 72 keV and gammas at 167 keV, and has a half-life of 73 h. Regulations allow an intrinsic contamination up to 3-5%, which is mainly caused by Tl-200 (368 keV; 26 h) and by Tl-202 (439 keV; 12.2 days). Contra-intuitive to the low-level percentages in which these contaminants are present, their impact may be significant because of much higher gamma camera sensitivity for these high-energy photon emissions. Therefore, we investigate the effects of the contaminants in terms of detected fractions of photons in projections and contrast degradation in reconstructed images. Methods Acquisitions of a digital thorax phantom filled with thallous (Tl-201) chloride were simulated with a validated Monte Carlo tool, thereby, modelling 1% of contamination by Tl-200 and Tl-202 each. In addition, measurements of a thorax phantom on a dual-headed gamma camera were performed. The product used was contaminated by 0.17% of Tl-200 and 0.24% of Tl-202 at activity reference time (ART). This ART is specified by the manufacturer, thereby, accounting for the difference in half-lives of Tl-201 and its contaminants. These measurements were repeated at different dates associated with various contamination levels. Results Simulations showed that, with 1% of Tl-200 and Tl-202, the total contamination in the 72 keV window can rise up to one out of three detected photons. For the 167keV window, the contamination is even more pronounced: more than four out of five detections in this photopeak window originate from contaminants. Measurements indicate that cold lesion contrast in myocardial perfusion SPECT imaging is at maximum close to ART. In addition to a higher noise level, relative contrast decreases 15% 2 days early to ART, which is explained by an increase in Tl-200 contamination. After ART, contrast decreased by 16% when the Tl-202 contamination increased to the maximal allowed limit. Conclusion Contra-intuitive to the low-level percentages in which they are typically present, penetration and downscatter of high-energy photons from Tl-200 and Tl-202 significantly contribute to thallous (Tl-201) chloride images, thereby, reducing contrast and adding noise. These findings may prompt for improved production methods, for updated policies with regard to timing of usage, and they also render the usefulness of adding the high photopeak window (167 keV) questionable. A window-based correction method for this contamination is advisable.