Photon-Counting Detector CT for Lenticulostriate Arteries: Comparison with Energy-Integrating Detector CT and Reconstruction Mode Optimization.

Abstract

Background and purpose: This study aimed to conduct a preliminary comparison of the visualization of lenticulostriate arteries (LSAs) in head and neck photon-counting detector computed tomography angiography (PCD-CTA) with four reconstruction modes.

Materials and methods: This retrospective study included patients who underwent head and neck CTA with PCD-CT between January 2024 and May 2024 and those who underwent CTA with energy-integrating detector CT (EID-CT) between September 2023 and May 2024. The images were reconstructed in four modes for the PCD-CT group and one mode for the EID-CT group. The image quality was evaluated, and the LSAs were visualized using subjective evaluations and quantitative assessments. The Friedman test was used to compare the characteristics of the four reconstruction modes.

Results: The PCD-CT group comprised 56 patients (average age 65 ± 14 years; 28 males), and the EID-CT group comprised 28 patients (average age, 64 ± 12 years; 14 males). In the PCD-CT group, the 0.2-mm-slice-thickness threshold 3 dimension (T3D) mode displayed the highest counts of LSAs (right side: 4.00 [range, 3.00-4.00]; left side: 4.00 [range, 3.00-4.00]) and the longest LSAs (right side: 23.33 ± 4.86 mm; left side: 24.18 ± 4.51 mm). The 0.4-mm-slice-thickness vascular spectral postprocessing (VSPP) mode obtained higher subjective evaluation scores. Furthermore, the PCD-CT group exhibited a higher count of LSAs, longer LSA lengths, and superior image quality than the EID-CT group.

Conclusions: PCD-CT achieves better visualization of LSAs and image quality than EID-CT and therefore constitutes an important method for observing LSAs in the future.

Abbreviations: PCD-CTA＝ photon-counting detector computed tomography angiography; LSAs＝ lenticulostriate arteries; EID-CT＝ energy-integrating detector CT; T3D＝ threshold 3 dimension; VSPP＝ vascular spectral postprocessing; CSVD＝ progression of cerebral small vessel disease; ICAD＝ intracranial atherosclerotic disease; UHR＝ ultra-high-resolution; QIR＝ quantum iterative reconstruction; ROI＝ region of interest; HU＝ hounsfield units; SNR＝ signal-to-noise ratio; CNR＝ contrast-to-noise ratio; FWHM＝ full width of half maximum; MIP＝ maximum intensity projection; CPR＝ curved planar reformation; ICC＝ intraclass correlation coefficient.