Estimation of bone strength at the millimeter and micrometer scales could aid in strength inference [1].
Trabecular bone consists of a network of oriented trabeculae that are typically spaced approximately 200 μm apart and measure approximately 80-100 μm [2]. Clinical CT scanners are limited in their ability to assess bone microstructure due to their detector size, which ranges from 0.5 to 0.625 mm at the isocenter. Therefore, detailed morphological parameters such as trabecular separation or thickness cannot be measured with sufficient accuracy and are overestimated [3, 4]. High-resolution (HR) quantitative peripheral CT can determine the geometric parameters of peripheral bone with a limiting resolution of 61-82 μm, although it is limited to the peripheral skeleton [4]. Medical micro-CT, considered the gold standard, also offers a very high spatial resolution (around 70 μm) but is only applicable to in vivo imaging of small animals [5].
Two new imaging techniques are commercially available: photon-counting detector CT (PCD-CT) and ultra-high resolution (UHR) CT . PCD-CT has been used to analyze trabecular bone in phantom and anatomical studies, has a spatial resolution equivalent to HR peripheral quantitative CT, and has demonstrated good correlation with micro-CT [3, 4, 9, 10]. Additionally, PCD-CT provides a lower radiation dose compared to conventional CT [10]. UHR-CT has also been compared with micro-CT in ex vivo phantom and anatomical studies, correlating well for cancellous bone parameters [11, 12]. The benefit of large-size matrices was also highlighted, as some parameters were identical between UHR-CT and micro-CT with a 20482 matrix size in tiger vertebrae [11, 12]. In addition, the combination of UHR-CT with Deep Learning Reconstruction (DLR) algorithms has been shown to improve the representation of cortical bone vascularization [8].
Because quantitative metrics of bone morphometry are useful indices for assessing overall bone quality, in this study we analyzed trabecular bone parameters using UHR-CT with and without DLR and at different matrix sizes and compared them with HR-CT using micro-CT as a reference in ex vivo specimens. The hypothesis was that UHR-CT with the larger matrix size would provide more accurate trabecular bone analysis than conventional CT, but less than micro-CT. This information may be useful for preclinical ground-truth imaging of bone microarchitecture to determine the usefulness of UHR-CT in clinical practice.