guest ::
| Home > Publications database > Assessment of trabecular microstructure using photon-counting CT: Influence of dose, spatial resolution and reconstruction algorithms. |
| Home > Publications database > Assessment of trabecular microstructure using photon-counting CT: Influence of dose, spatial resolution and reconstruction algorithms. |
| Journal Article | DKFZ-2026-02146 |
; ; ; ; ; ; ; ; ; ; ; ;
2026
Elsevier Science
Amsterdam [u.a.]
Abstract: Photon-counting CT (PCCT) for quantitative microstructural assessments of the central skeleton remains underexplored. This study evaluated the effects of radiation dose, spatial resolution, noise and reconstruction algorithm on bone mineral density (BMD), bone volume fraction (BV/TV), and trabecular separation (Tb.Sp).Seven excised human vertebral bodies embedded in polymethylmethacrylate were scanned using a clinical PCCT scanner (Naeotom Alpha, Siemens Healthineers): natively at (20 mGy CTDI32 cm) and within a thorax phantom at (6, 9, 12 and 15 mGy). Four repeated scans (15 mGy) served as reference. Filtered back projection (FBP) and iterative (QIR3) reconstructions were used with Br56u, Br76u and Br89u kernels. Bone parameters were estimated from cancellous volumes. Differences were tested with Wilcoxon signed-rank tests. Radiation dose impact was assessed using Bland-Altman analysis and Lin's concordance coefficient (rccc).Median (IQR) ranges for BMD were 177.5 (55.1)-191.9 (57.5) mgHA/cm3, for BV/TV 0.25 (0.20)-0.47 (0.05), and Tb.Sp 0.29 (0.02)-1.50 (1.64) mm. FBP consistently yielded higher BMD and BV/TV than QIR3, albeit non-significant after statistical correction. BMD differences between algorithms were 6.6 mgHA/cm3 at 6.0 mGy Br89u, and 0.1 mgHA/cm3 for Br76u at 20 mGy. Noise was higher in FBP (range: 13-470 HU) vs QIR3 (6-124 HU). The mean absolute (%) differences to the reference across dose levels were best for Br76u/QIR3 (BMD: 0.83 mgHA/cm3 (0.5%); BV/TV: 0.01 (2.6%) and Tb.Sp: 0.79 mm (18%)). Best robustness against radiation dose changes was observed with Br56u/QIR3 and Br76u/QIR3.Br56u showed enlarged Tb.Sp levels despite excellent agreement, while Br89u showed increased noise levels and instability. Br76u/QIR3 showed the best balance between microstructural accuracy, image noise and stability at clinically diagnostic dose levels, for example 15 mGy CTDIvol.
Keyword(s): Filtered back projection ; Iterative reconstruction ; Photon-counting CT ; Quantitative bone analysis ; Trabecular bone
; BIOSIS Previews ; Biological Abstracts ; Clarivate Analytics Master Journal List ; Current Contents - Clinical Medicine ; Current Contents - Life Sciences ; Ebsco Academic Search ; Essential Science Indicators ; IF < 5 ; JCR ; Nationallizenz
; SCOPUS ; Science Citation Index Expanded ; Web of Science Core Collection
|
The record appears in these collections: |
