Individualized assessment of osteoporotic fracture risk at the spine using ultra-low-dose MDCT imaging techniques and non-dedicated routine MDCT exams

Background: Osteoporosis is a skeletal disorder characterized by compromised bone strength, increasing fracture risk. Osteoporotic fractures, especially at the spine and hip, reduce quality of life and increase morbidity and mortality, making osteoporosis a significant public health issue. Current screening technologies, namely Dual Energy X-ray Absorptiometry (DXA)-based Bone Mineral Density (BMD) measurements and T-scores, have limitations in fracture risk prediction. Opportunistic volumetric vertebral BMD assessment and Finite Element Modelling (FEM) using Multi-Detector Computed Tomography (MDCT) perform comparably or better than DXA-based T-scores in predicting fractures. Our research was focused on the optimization of screening techniques and had two primary objectives: 1. Enhancing Imaging Techniques for Bone Strength Prediction We evaluated advanced CT acquisition models, including virtual sparse sampling and dual-layer spectral CT, enabling ultra-low dose BMD and FEM analysis. Results indicated a 50–90% radiation dose reduction while maintaining accurate vertebral bone strength prediction. Spectral-detector-based x-ray absorptiometry (SDXA) successfully differentiated patients with and without fractures, supporting its potential for opportunistic osteoporosis screening. Additionally, iodine-corrected BMD measurements from contrast-enhanced spectral CT demonstrated high accuracy, making opportunistic BMD assessment feasible. 2. Automating Opportunistic Osteoporosis Screening A fully automated framework was developed for MDCT-based osteoporosis screening, integrating vertebral segmentation, trabecular and integral volumetric BMD extraction, and FEM-based failure load assessment. Automated BMD measurements showed strong agreement with standard QCT and outperformed DXA in fracture prediction. Threshold values for fracture risk assessment were established, and prediction models based on extracted BMD surpassed vertebral fracture count-based models. Summary: Our studies highlight the feasibility of ultra-low dose CT for bone strength prediction and the potential of deep learning for opportunistic osteoporosis screening in clinical MDCT data.     «

Background: Osteoporosis is a skeletal disorder characterized by compromised bone strength, increasing fracture risk. Osteoporotic fractures, especially at the spine and hip, reduce quality of life and increase morbidity and mortality, making osteoporosis a significant public health issue. Current screening technologies, namely Dual Energy X-ray Absorptiometry (DXA)-based Bone Mineral Density (BMD) measurements and T-scores, have limitations in fracture risk prediction. Opportunistic volumetric...     »