Improving the image quality of short-time bone SPECT using cadmium-zinc-telluride detectors with SwiftScan.

Abstract

Objectives: This study aimed to evaluate the quality and associated quantitative values of bone single-photon emission computed tomography (SPECT) with and without SwiftScan using a semiconductor camera equipped with a cadmium-zinc-telluride detector.

Methods: Ten patients with bone metastases from prostate cancer who underwent list-mode SPECT/computed tomography using a whole-body semiconductor camera participated in this study. A total of 130 metastatic lesions from 10 patients were analyzed. Standard SPECT images were obtained approximately 3 h later, and the images were constructed with and without SwiftScan.

Results: The visual assessment of 3-dimensional maximum intensity projection images showed that when an image quality score of 4 (good) or better was considered clinically acceptable, it was maintained at 4 or better in the 75% and 50% scans with SwiftScan, whereas only the 75% scan was considered acceptable without SwiftScan. The intraclass correlation coefficient was 0.952 at 5% for the standard time without SwiftScan and 0.990 with SwiftScan. The maximum standardized uptake value (SUV_max) changes were 0 to 9.5 (median 1.1) at 75%, 0.1 to 11.5 (1.65) at 50%, 0 to 15.7 (2.1) at 25%, 0.1 to 33.2 (4.2) at 10%, 0.2 to 8.9 (5.65) at 5% without SwiftScan. On the contrary, the SUV_max changes in absolute value were 0 to 5.4 (median 0.8) at 75%, 0 to 6.5 (1.4) at 50%, 0 to 19.1 (1.7) at 25%, 0 to 24.2 (2.8) at 10%, 0 to 29.9 (2.6) at 5% with SwiftScan. The contrast-to-noise ratios (CNR) were 95.3 at 75%, 88.3 at 50%, 69.2 at 25%, 45.7 at 10%, and 31.6 at 5% without SwiftScan, and 96.9, 91.7, 78.0, 71.6, and 62.0, respectively, using SwiftScan.

Conclusion: With the use of SwiftScan, a 50% reduction in acquisition time was considered acceptable for image quality with reproducible quantitative indices such as SUV_max and CNR.