Capabilities of multi-pinhole SPECT with two stationary detectors for in vivo rat imaging

• Published in: Scientific reports Vol. 10; no. 1; p. 18616
• Main Authors: Janssen, Jan P., Hoffmann, Jan V., Kanno, Takayuki, Nose, Naoko, Grunz, Jan-Peter, Onoguchi, Masahisa, Chen, Xinyu, Lapa, Constantin, Buck, Andreas K., Higuchi, Takahiro
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 29.10.2020; Nature Publishing Group
• Subjects: 692/308; 692/308/2778; 692/308/575; 692/53; 692/53/2421; Animals; Bone and Bones - diagnostic imaging; Computed tomography; Equipment Design - instrumentation; Equipment Design - methods; Heart; Heart Ventricles - diagnostic imaging; Humanities and Social Sciences; Image Processing, Computer-Assisted - instrumentation; Image Processing, Computer-Assisted - methods; Mice; multidisciplinary; Performance assessment; Phantoms, Imaging; Radioactive tracers; Rats; Science; Science (multidisciplinary); Sensors; Single photon emission computed tomography; Spatial discrimination; Technetium - administration & dosage; Technetium Tc 99m Sestamibi - administration & dosage; Tomography, Emission-Computed, Single-Photon - instrumentation; Tomography, Emission-Computed, Single-Photon - methods; Ventricle
• DOI: 10.1038/s41598-020-75696-0
• ISSN: 2045-2322

We aimed to investigate the image quality of the U-SPECT5/CT E-Class a micro single-photon emission computed tomography (SPECT) system with two large stationary detectors for visualization of rat hearts and bones using clinically available 99m Tc-labelled tracers. Sensitivity, spatial resolution, uniformity and contrast-to-noise ratio (CNR) of the small-animal SPECT scanner were investigated in phantom studies using an ultra-high-resolution rat and mouse multi-pinhole collimator (UHR-RM). Point source, hot-rod, and uniform phantoms with 99m Tc-solution were scanned for high-count performance assessment and count levels equal to animal scans, respectively. Reconstruction was performed using the similarity-regulated ordered-subsets expectation maximization (SROSEM) algorithm with Gaussian smoothing. Rats were injected with ~ 100 MBq [ 99m Tc]Tc-MIBI or ~ 150 MBq [ 99m Tc]Tc-HMDP and received multi-frame micro-SPECT imaging after tracer distribution. Animal scans were reconstructed for three different acquisition times and post-processed with different sized Gaussian filters. Following reconstruction, CNR was calculated and image quality evaluated by three independent readers on a five-point scale from 1 = “very poor” to 5 = “very good”. Point source sensitivity was 567 cps/MBq and radioactive rods as small as 1.2 mm were resolved with the UHR-RM collimator. Collimator-dependent uniformity was 55.5%. Phantom CNR improved with increasing rod size, filter size and activity concentration. Left ventricle and bone structures were successfully visualized in rat experiments. Image quality was strongly affected by the extent of post-filtering, whereas scan time did not have substantial influence on visual assessment. Good image quality was achieved for resolution range greater than 1.8 mm in bone and 2.8 mm in heart. The recently introduced small animal SPECT system with two stationary detectors and UHR-RM collimator is capable to provide excellent image quality in heart and bone scans in a rat using standardized reconstruction parameters and appropriate post-filtering. However, there are still challenges in achieving maximum system resolution in the sub-millimeter range with in vivo settings under limited injection dose and acquisition time.