Quantitative Evaluation of Hip Impingement Utilizing Computed Tomography Measurements

CT in musculoskeletal imaging: still helpful and for what?

Computed tomography (CT) is a common modality employed for musculoskeletal imaging. Conventional CT techniques are useful for the assessment of trauma in detection, characterization and surgical planning of complex fractures. CT arthrography can depict internal derangement lesions and impact medical decision making of orthopedic providers. In oncology, CT can have a role in the characterization of bone tumors and may elucidate soft tissue mineralization patterns. Several advances in CT technology have led to a variety of acquisition techniques with distinct clinical applications. These include four-dimensional CT, which allows examination of joints during motion; cone-beam CT, which allows examination during physiological weight-bearing conditions; dual-energy CT, which allows material decomposition useful in musculoskeletal deposition disorders (e.g., gout) and bone marrow edema detection; and photon-counting CT, which provides increased spatial resolution, decreased radiation, and material decomposition compared to standard multi-detector CT systems due to its ability to directly translate X-ray photon energies into electrical signals. Advanced acquisition techniques provide higher spatial resolution scans capable of enhanced bony microarchitecture and bone mineral density assessment. Together, these CT acquisition techniques will continue to play a substantial role in the practices of orthopedics, rheumatology, metabolic bone, oncology, and interventional radiology.

Measurement of acetabular wall indices: comparison between CT and plain radiography

The purpose of this study was to compare measurements of anterior wall index (AWI) and posterior wall index (PWI) on computed tomography (CT) to those on radiographs (XR). A consecutive cohort of 33 patients (45 hips total) being evaluated for hip pain with both XR and CT was examined. Preoperative measurements of AWI and PWI were performed utilizing supine anteroposterior pelvic XR and coronal and swiss axial CT scans by two independent raters. Mean differences between XR and CT measurements were compared, and agreement between measurements was assessed using the concordance correlation coefficient (r_c) and Bland–Altman analysis. A total of 39 hips in 28 patients were analyzed. The mean patient age was 31.1 ± 9.0 years, and 50% were female. Mean AWI and PWI on XR was 0.50 ± 0.14 and 0.91 ± 0.12, respectively. Measured values of AWI were consistently larger (0.08 ± 0.10, P < 0.01) on XR compared with both coronal and swiss axial CT, with moderate agreement between XR and CT measurements (r_c = 0.68–0.70). Measured values of PWI were consistently smaller (0.15 ± 0.12, P < 0.05) on XR compared with both coronal and swiss axial CT, with poor agreement between XR and CT measurements (r_c = 0.37–0.45). Measured values of acetabular wall indices on XR were consistently larger for AWI and smaller for PWI relative to CT. Agreement between XR and CT measures of the indices were moderate to poor. This highlights the need for standardization of XR- and CT-based measurements to improve assessment of acetabular coverage and subsequent clinical decision-making.

Semi-supervised labelling of the femur in a whole-body post-mortem CT database using deep learning

A deep learning pipeline was developed and used to localize and classify a variety of implants in the femur contained in whole-body post-mortem computed tomography (PMCT) scans. The results provide a proof-of-principle approach for labelling content not described in medical/autopsy reports. The pipeline, which incorporated residual networks and an autoencoder, was trained and tested using n = 450 full-body PMCT scans. For the localization component, Dice scores of 0.99, 0.96, and 0.98 and mean absolute errors of 3.2, 7.1, and 4.2 mm were obtained in the axial, coronal, and sagittal views, respectively. A regression analysis found the orientation of the implant to the scanner axis and also the relative positioning of extremities to be statistically significant factors. For the classification component, test cases were properly labelled as nail (N⁺), hip replacement (H⁺), knee replacement (K⁺) or without-implant (I^-) with an accuracy >97%. The recall for I^- and H⁺ cases was 1.00, but fell to 0.82 and 0.65 for cases with K⁺ and N⁺. This semi-automatic approach provides a generalized structure for image-based labelling of features, without requiring time-consuming segmentation.