Feasibility of ultra-high-speed acquisition in xSPECT bone algorithm: a phantom study with advanced bone SPECT-specific phantom

• Published in: Annals of nuclear medicine 2022-02, Vol.36 (2), p.183-190
• Main Authors: Ichikawa, Hajime, Miyaji, Noriaki, Onoguchi, Masahisa, Shibutani, Takayuki, Nagaki, Akio, Kato, Toyohiro, Shimada, Hideki
• Format: Article
• Language: English
• Subjects: Algorithms; Bones; Coefficient of variation; Computed tomography; Customization; Feasibility; High speed; Image acquisition; Image quality; Image reconstruction; Image resolution; Imaging; Lung cancer; Medical imaging; Medicine; Medicine & Public Health; Nuclear Medicine; Original Article; Patients; Photon emission; Radioactivity; Radiology; Single photon emission computed tomography; Thorax; Tomography; Vertebrae
• ISSN: 0914-7187; 1864-6433
• DOI: 10.1007/s12149-021-01689-2

Objective Although xSPECT Bone (xB) provides quantitative single-photon emission computed tomography (SPECT) high-resolution images, patients’ burden remains high due to long acquisition time; therefore, this study aimed to investigate the feasibility of shortening the xB acquisition time using a custom-designed phantom. Methods A custom-designed xSPECT bone-specific (xSB) phantom with simulated cortical and spongious bones was developed based on the thoracic bone phantom. Both standard- and ultra-high-speed (UHS) xB acquisitions were performed in a male patient with lung cancer. In this phantom study, SPECT was acquired for 3, 6, 9, 12, and 30 min. The clinical SPECT acquisition time per rotation was 9 and 3 min for standard and UHS, respectively. SPECT images were reconstructed using ordered subset expectation maximization with three-dimensional resolution recovery (Flash3D; F3D) and xB algorithms. Quantitative SPECT value (QSV) and coefficient of variation (CV) were measured using the volume of interests (VOIs) placed at the center of the vertebral body and hot sphere. A linear profile was plotted on the spinous process at the center of the xSB phantom; then, the full width at half maximum (FWHM) was measured. The standardized uptake value (SUV) and standard deviation from the first thoracic to the fifth lumbar vertebrae in clinical standard- and UHS-xB images were measured using a 1-cm 3 VOI. Results The QSV of F3D images was underestimated even in large regions, whereas those of xB images were close to actual radioactivity concentration. The CV was similar or lower for xB images than that for F3D images but was not decreased with increasing acquisition time for both reconstruction images. The FWHM of xB images was lower than those of F3D images at all acquisition times. The mean SUV values from the first thoracic to fifth lumbar vertebrae for standard- and UHS-xB images were 6.73 ± 0.64 and 6.19 ± 0.87, respectively, showing a strong positive correlation. Conclusions Results of this phantom study suggest that xB imaging can be obtained in only one-third of the acquisition time without compromising the image quality. The SUV of UHS-xB images can be similar to that of standard-xB images in terms of clinical interpretation.