Bone disease can have a devastating impact on an individual’s life quality and expectancy. Bone fracture is characterized as the most prevalent musculoskeletal disorder that necessitates hospitalization and it is a common consequence of osteoporosis. Bone microarchitecture is highly affected by bone diseases, such as osteoporosis and osteoarthritis, through the alteration of its structure and integrity. It has been proven that the visualization and quantification of bone structure on the micro-scale is a crucial requirement for the early detection of bone fractures and disorders. Photon Counting Detector Computed Tomography (PCD-CT) is a novel clinical imaging system that provides the potential of high spatial resolution scans while using low radiation dose. However, its inherent advantages over traditional CT scanners have not been widely investigated yet. The aim of this study was to evaluate the efficiency of PCD-CT in quantifying features of bone microstructures like bone volume fraction, trabecular thickness, separation and number, in various radiation dose
levels. The capabilities of PCD-CT in bone morphological analysis were further investigated through the comparison against the gold-standard techniques of Micro-Computed Tomography (micro-CT) and High Resolution Peripheral Quantitative Computed Tomography (HR-pQCT). Four pairs of cadaver bones acquired from different anatomical locations (radius, tibia, vertebrae, femoral head) of the same individual were scanned with three CT scanners, PCD-CT, micro-CT and HR-pQCT, using micro-CT as reference standard. The bones of vertebrae and femoral head were scanned within a human body-simulating phantom, as well. The average measurement of the four bone parameters was calculated using an automated volume-based fitting and a voxel-counting algorithm. The image quality of PCD-CT was also assessed by estimating the trabecular sharpness and contrast to noise ratio, through an algorithm which analyzed the histogram values of localized bone and fat regions. The comparison between PCD-CT and micro-CT/HR-pQCT extracted data was realized through the
statistical analysis processes of Bland & Altman and linear regression analysis. The results of this study indicated that the ratio of bone volume to total volume as well as the thickness and separation of trabeculae were overestimated by PCD-CT in comparison to the micro-CT, in the bone scans of all cadaver bones. In the case of bones within the phantom, the thickness and separation of trabeculae were overestimated but the bone volume to total volume was underestimated compared to micro-CT. The number of trabeculae was consistently underestimated in both cases in which the bones were scanned within the phantom or not. In the comparison against HR-pQCT, the bone volume to total volume ratio, number and thickness of trabeculae were overestimated by PCD-CT whereas the trabecular separation was underestimated. Through this study, it was indicated that the correlation between PCD-CT and micro-CT or HR-pQCT extracted bone parameters improves as higher doses are used during PCD-CT scanning. However, the agreement between the extracted bone parameters does not necessarily increase in the same way. Overall, PCD-CT featured sufficient image resolution and sharpness to evaluate the morphology of the bone, in both low and high radiation doses, with
the highest image quality being achieved through the high dose values. These outcomes suggest that PCD-CT could hold great promise for analysing the bone microstructure in the peripheral and central human skeleton, revolutionizing the diagnosis and treatment of bone pathologies.

Image registration combines data from multiple imaging modalities or time points by aligning images. Existing 2D-3D registration methods often require manual intervention, introducing variability and increasing processing time, and rely on Computed Tomography (CT) imaging, which exposes patients to radiation. This study aims to develop an automated registration pipeline that ensures accurate registration between 3D Magnetic Resonance Imaging (MRI) and 2D fluoroscopic images by preprocessing techniques.

This study involved two registration experiments to evaluate the pipeline. The DiffDRR package was used with normalized cross-correlation (NCC) as similarity metric and gradient-based Adam optimizer for optimization. The first experiment validated the accuracy of registration between 3D MRI models (solid bone, empty shell, and filled shell) and 2D CT-derived digitally reconstructed radiographs (DRRs) with known rotations and translations as reference standard. The second experiment aimed to determine the most effective fluoroscopy preprocessing technique for aligning 3D MRI models with 2D fluoroscopic images. Preprocessing techniques for the fluoroscopic images included contrast enhancement via thresholding and normalization, and region of interest selection or segmentation.

Validation results showed smaller rotational and translational errors for both shell models compared to the solid bone model. In-plane errors for the shell models were 0.1-3.3mm and out-of-plane 15.8-55.0mm, while solid bone reported in-plane errors of 1.0-4.7mm and out-of-plane of 68.3-282.7mm. Registration between empty shell models and differently preprocessed fluoroscopic images showed that segmentation of the bones in fluoroscopic images yielded the highest NCC values. However, visual assessment revealed inconsistencies in plausible bone positioning.

This automated 2D-3D registration method shows potential for registration between 3D MRI and 2D fluoroscopic images of the knee joint. The current method requires further refinement, particularly in integrating soft tissue data, possibly via synthetic CTs, and adopting more anatomically sensitive metrics, such as mutual information or gradient correlation.