Computation of the effective atomic number and mass density from virtual monoenergetic photon-counting CT reconstructions

The aim of this work was to develop and test an image-domain-based method for calculating mass density ( ρ ) and effective atomic number ( Z eff ) maps from virtual monoenergetic reconstructions generated with the clinically approved photon-counting computed tomography (CT) system Naeotom Alpha (Siemens Healthineers, Erlangen, Germany). For this purpose, we developed an algorithm that utilizes virtual monoenergetic images (VMI) at different energy levels to calculate ρ - and Z eff -maps. This approach is computationally efficient because preprocessing steps for scanner geometry, beam hardening, beam quality, and spectral detector efficiency are not required, as these corrections are already incorporated in the manufacturer-generated VMI data. In the framework of a phantom study, we demonstrated high accuracy of the computed values compared with theoretically expected reference values. The algorithm can be used for analyzing biological materials with mass densities up to 2.0 g/cm 3 . Within this range, we obtained an average absolute percentage deviation of 1.0% for ρ -measurements. For Z eff -measurements, the average absolute percentage deviation was 0.84%. These small deviations of the experimental data from the expected values are in the same order of magnitude as deviations obtained with more complex methods reported in the literature. We conclude that virtual monoenergetic images from photon-counting CT can be used for the calculation of ρ - and Z eff -maps of biological materials with mass densities up to 2.0 g/cm 3 , which is sufficient for measurements in the human tissues. The proposed method is efficient, transparent, and directly applicable to clinically available image-domain data without requiring raw projection data or scanner-specific preprocessing.