Energy-resolved attenuation behaviour on dual-source photon-counting CT: Implications for quantitative assessment of femoral bone marrow in multiple myeloma

Rationale and objectives Quantitative CT assessment of multiple myeloma relies on threshold-based differentiation between physiological tissue and myeloma infiltration, yet bone marrow attenuation may be influenced by photon energy. The aim of this study was to evaluate the energy-dependent attenuation behaviour of femoral bone marrow on virtual monoenergetic images derived from dual-source photon-counting CT (PCD-CT) and its implications for quantitative analysis. Methods This retrospective single-centre study included 103 patients who underwent unenhanced ultra-high-resolution whole-body PCD-CT for multiple myeloma assessment. VMI were reconstructed at 40–190 keV. Attenuation was measured in femoral bone marrow, adjacent muscle, and subcutaneous fat. Contrast-to-noise ratios were calculated and differences across energy levels were analysed using Friedman tests with Bonferroni correction. Results Femoral bone marrow attenuation increased with higher keV from −83.4 HU at 40 keV to −58.7 HU at 190 keV, while fat attenuation increased more markedly and muscle attenuation decreased (all p < 0.001). As a result, the attenuation difference between marrow and fat declined substantially from 99.9 HU at 40 keV to 26.9 HU at 190 keV (p < 0.001). Although image noise decreased toward higher keV, CNR declined from 8.4 at 40 keV to 4.1 at 190 keV (p < 0.001). Conclusions Bone marrow attenuation on dual-source PCD-CT is energy-dependent and should not be interpreted independently of the selected monoenergetic level. Quantitative assessment requires contextualization relative to surrounding tissues rather than reliance on fixed attenuation thresholds.