Cadmium zinc telluride (CZT) solid-state detectors have been recently introduced in myocardial perfusion imaging. However, they had not been yet validated with thallium-201. This study compares the clinical performances of the CZT ultrafast camera GE DNM 530c with a conventional SPECT camera (CC) using thallium-201. We prospectively studied with thallium-201 a total of 153 consecutive patients referred for myocardial perfusion imaging at exercise (3-4 mCi) then redistribution (with 1 mCi reinjection). Sequential acquisitions were performed first with a conventional dual-head tomographic Anger camera (CC) in 10 to 15 minutes and then with a CZT camera (CZT) in 5 minutes, in prone position. In all, 9 patients were excluded: 1 for mispositioning, 3 for camera failure, 3 for delayed acquisition after exercise, 1 for nonacceptance of redistribution, 1 for motion. Acquisition was more comfortable with CZT for all patients. Global counts rate was higher with CZT than with CC (3.6±0.57 KCts/s vs. 1.14±0.16). CZT has a 5-fold increased myocardial counts rate compared with CC (448±69 Kcts in 5 minutes vs. 209±40 Kcts in 12.5±1.8 minutes). Quality of CZT images was considered as better in 40%, equal in 56%, and worse in 4% of cases; we found less artifacts with CZT; diagnostic conclusions were the same in 140 of 144 cases (97%); discordances were 2 artifacts with CC and 2 small ischemia (less than 2 segments) missed by CZT. This new dedicated cardiac CZT camera allows with thallium-201 five minutes acquisitions with an increased image quality and a reliable diagnosis quality.

Thallium-201 is efficient for myocardial perfusion imaging, but leads to relatively high radiation exposure in patients. The purpose of this study was to compare the efficiency of low-dose thallium-201 imaging with cadmium-zinc-telluride (CZT) cameras with regular-dose thallium-201 imaging with conventional cameras. We prospectively studied 137 consecutive patients referred for stress myocardial perfusion imaging who had previously had a myocardial single-photon emission computed tomography with thallium-201. We injected at stress a low dose of thallium-201 (1.1 MBq/kg, 28 µCi/kg), performed a 5-7 min scan with a CZT camera (GE DNM 530c), and assessed redistribution imaging when the initial images were abnormal. We compared the CZT scan with the conventional dual-head tomographic camera scan taken previously with a regular dose of thallium-201. The average delay between both scans was 22 months. The stress dose was 88 ± 16 (2.38 ± 0.43 mCi) versus 125 ± 13 MBq (3.38 ± 0.34 mCi; a 30% reduction). The time for camera acquisition was 6 versus 13 min (a 54% reduction). The myocardial counts were increased two-fold with CZT (mean: 446 Kcounts). The quality of CZT images was better in 69% of the cases. There were 59 artifacts with conventional cameras and 29 artifacts with CZT (P<0.01). The diagnostic agreement was calculated in patients without clinical or angiographic changes between both scans (115 patients) and was high (97%). The effective dose at stress was less than 12 mSv. High myocardial counting allowed for further decrease in the injected activity, leading to an effective dose as low as 8 mSv. With reduced activities of thallium-201 and low effective doses, the CZT camera provides reliable, high-quality imaging.

(1) Background: The CT-based attenuation correction of SPECT images is essential for obtaining accurate quantitative images in cardiovascular imaging. However, there are still many SPECT cameras without associated CT scanners throughout the world, especially in developing countries. Performing additional CT scans implies troublesome planning logistics and larger radiation doses for patients, making it a suboptimal solution. Deep learning (DL) offers a revolutionary way to generate complementary images for individual patients at a large scale. Hence, we aimed to generate linear attenuation coefficient maps from SPECT emission images reconstructed without attenuation correction using deep learning. (2) Methods: A total of 384 SPECT myocardial perfusion studies that used ^99mTc-sestamibi were included. A DL model based on a 2D U-Net architecture was trained using information from 312 patients. The quality of the generated synthetic attenuation correction maps (ACMs) and reconstructed emission values were evaluated using three metrics and compared to standard-of-care data using Bland–Altman plots. Finally, a quantitative evaluation of myocardial uptake was performed, followed by a semi-quantitative evaluation of myocardial perfusion. (3) Results: In a test set of 66 test patients, the ACM quality metrics were MSSIM = 0.97 ± 0.001 and NMAE = 3.08 ± 1.26 (%), and the reconstructed emission quality metrics were MSSIM = 0.99 ± 0.003 and NMAE = 0.23 ± 0.13 (%). The 95% limits of agreement (LoAs) at the voxel level for reconstructed SPECT images were: [−9.04; 9.00]%, and for the segment level, they were [−11; 10]%. The 95% LoAs for the Summed Stress Score values between the images reconstructed were [−2.8, 3.0]. When global perfusion scores were assessed, only 2 out of 66 patients showed changes in perfusion categories. (4) Conclusion: Deep learning can generate accurate attenuation correction maps from non-attenuation-corrected cardiac SPECT images. These high-quality attenuation maps are suitable for attenuation correction in myocardial perfusion SPECT imaging and could obviate the need for additional imaging in standalone SPECT scanners.