Strength and strain distributions obtained from digital wrist tomosynthesis discriminate patients with and without a history of fragility fracture.

Bone fractures due to osteoporosis are a significant problem. Limited accuracy of standard bone mineral density (BMD) for fracture risk assessment, combined with low adherence to bone health screening precludes identification of those at risk of fracture. Because of the wide availability of digital breast tomosynthesis (DBT) imaging, bone screening using a DBT scanner at the time of breast screening has been proposed. Earlier studies have shown that BMD, microstructure, and stiffness of the distal radius can be calculated using digital tomosynthesis imaging of the wrist (DWT). However, strength and stress/strain parameters, which are more relevant to structural failure, and have the potential to enhance the utility of DWT, were not examined previously. Therefore, this study aimed to examine the ability of DWT to discriminate patients with and without fragility fracture using DWT based finite element (DWT-FE) derived strength and stress/strain distribution properties, and to determine in vivo repeatability of these biomechanical properties. Twenty-two postmenopausal women with any fragility fracture (included spine, hip, distal radius, humerus and tibia fractures) and 68 without were recruited. Each participant's nondominant arm (dominant arm if history of fracture in the nondominant arm) was scanned with DWT and compressive loading was simulated using FE modeling. Six additional patients were DWT-scanned thrice, with repositioning, to determine the repeatability of the study variables. Age and T-score were not different between fracture and nonfracture groups (p > 0.1), but strength and stress/strain parameters were significant predictors of fracture status (AUC = 0.64-0.74). Standard deviation of tensile strain was the most discriminatory variable for fracture status (AUC = 0.74) and was independent from stiffness. Repeatability error of DWT biomechanical properties was 0.7 % to 5.8 %. This study demonstrated that DWT-FE based strength and standard deviation of tensile strain were reproducible and predict fracture status independent from BMD and stiffness. The results suggest that the accuracy of fracture risk screening can be improved in the highly accessible environment of mammographic imaging.