Thomas L. Nickolas | Searcharxiv

arXiv2026

Short-Term Precision and Least Significant Change of 3D-DXA Cortical and Trabecular Proximal Femur Measurements Across Hologic DXA Scanner Models

3D-DXA provides volumetric and compartment-specific hip measurements from standard DXA scans. However, for reliable longitudinal interpretation, establishing scanner-specific short-term precision data is essential. This study assessed the short-term precision of 3D-DXA-derived parameters using repeated acquisitions from five clinical Hologic DXA scanner datasets. Duplicate hip DXA acquisitions were collected at five clinical centers using two Horizon Wi scanners, two Horizon A scanners, and one Discovery W scanner. Each subject was scanned twice with complete repositioning between acquisitions. Conventional total hip and femoral neck aBMD were obtained using APEX software, and 3D-Shaper software was used to derive integral vBMD, trabecular vBMD, and cortical sBMD. Precision error was expressed as RMS-SD and RMS-CV, and LSC was calculated at the 95% confidence level. For total hip aBMD, absolute LSC values ranged from 0.018 to 0.037 g/cm$^2$. Femoral neck aBMD showed higher error, with LSC values ranging from 0.031 to 0.049 g/cm$^2$. For 3D-DXA parameters, absolute LSC values ranged from 9.955 to 17.859 mg/cm$^2$ for integral vBMD, 8.662 to 18.487 mg/cm$^2$ for trabecular vBMD, and 4.444 to 8.875 mg/cm$^2$ for cortical sBMD. Precision errors for 3D-Shaper parameters varied moderately across the five clinical datasets. Short-term precision of 3D-Shaper-derived measurements was broadly comparable with previously published precision data. The LSC values reported here can support interpretation of longitudinal changes in integral vBMD, trabecular vBMD, and cortical sBMD on Hologic scanners.

q-bio.QM↗

arXiv2026

Transformer-Based Multi-Region Segmentation and Radiomic Analysis of HR-pQCT Imaging for Osteoporosis Classification

Osteoporosis is a skeletal disease typically diagnosed using dual-energy X-ray absorptiometry (DXA), which quantifies areal bone mineral density but overlooks bone microarchitecture and surrounding soft tissues. High-resolution peripheral quantitative computed tomography (HR-pQCT) enables three-dimensional microstructural imaging with minimal radiation. However, current analysis pipelines largely focus on mineralized bone compartments, leaving much of the acquired image data underutilized. We introduce a fully automated framework for binary osteoporosis classification using radiomics features extracted from anatomically segmented HR-pQCT images. To our knowledge, this work is the first to leverage a transformer-based segmentation architecture, i.e., the SegFormer, for fully automated multi-region HR-pQCT analysis. The SegFormer model simultaneously delineated the cortical and trabecular bone of the tibia and fibula along with surrounding soft tissues and achieved a mean F1 score of 95.36%. Soft tissues were further subdivided into skin, myotendinous, and adipose regions through post-processing. From each region, 939 radiomic features were extracted and dimensionally reduced to train six machine learning classifiers on an independent dataset comprising 20,496 images from 122 HR-pQCT scans. The best image level performance was achieved using myotendinous tissue features, yielding an accuracy of 80.08% and an area under the receiver operating characteristic curve (AUROC) of 0.85, outperforming bone-based models. At the patient level, replacing standard biological, DXA, and HR-pQCT parameters with soft tissue radiomics improved AUROC from 0.792 to 0.875. These findings demonstrate that automated, multi-region HR-pQCT segmentation enables the extraction of clinically informative signals beyond bone alone, highlighting the importance of integrated tissue assessment for osteoporosis detection.

cs.CV↗