Accurately estimating whole bone strength is critical for identifying individuals that may benefit from prophylactic treatments aimed at reducing fracture risk. Strength is often estimated from stiffness, but it is not known whether the relationship between stiffness and strength varies with age and sex. Cadaveric proximal femurs (44 Male: 18-78 years; 40 Female: 24-95 years) and radial (36 Male: 18-89 years; 19 Female: 24-95 years) and femoral diaphyses (34 Male: 18-89 years; 19 Female: 24-95 years) were loaded to failure to evaluate how the stiffness-strength relationship varies with age and sex. Strength correlated significantly with stiffness at all sites and for both sexes, as expected. However, females exhibited significantly less strength for the proximal femur (58% difference, p<0.001). Multivariate regressions revealed that stiffness, age and PYD were significant negative independent predictors of strength for the proximal femur

Given prior work showing associations between remodeling and external bone size, we tested the hypothesis that wide bones would show a greater negative correlation between whole‐bone strength and age compared with narrow bones. Cadaveric male radii (n = 37 pairs, 18 to 89 years old) were evaluated biomechanically, and samples were sorted into narrow and wide subgroups using height‐adjusted robustness (total area/bone length). Strength was 54% greater (p < 0.0001) in wide compared with narrow radii for young adults (<40 years old). However, the greater strength of young‐adult wide radii was not observed for older wide radii, as the wide (R2 = 0.565, p = 0.001), but not narrow (R2 = 0.0004, p = 0.944) subgroup showed a significant negative correlation between strength and age. Significant positive correlations between age and robustness (R2 = 0.269, p = 0.048), cortical area (Ct.Ar; R2 = 0.356, p = 0.019), and the mineral/matrix ratio (MMR; R2 = 0.293, p = 0.037) were observed for narrow, but not wide radii (robustness: R2 = 0.015, p = 0.217; Ct.Ar: R2 = 0.095, p = 0.245; MMR: R2 = 0.086, p = 0.271). Porosity increased with age for the narrow (R2 = 0.556, p = 0.001) and wide (R2 = 0.321, p = 0.022) subgroups. The wide subgroup (p < 0.0001) showed a significantly greater elevation of a new measure called the Cortical Pore Score, which quantifies the cumulative effect of pore size and location, indicating that porosity had a more deleterious effect on strength for wide compared with narrow radii. Thus, the divergent strength–age regressions implied that narrow radii maintained a low strength with aging by increasing external size and mineral content to mechanically offset increases in porosity. In contrast, the significant negative strength–age correlation for wide radii implied that the deleterious effect of greater porosity further from the centroid was not offset by changes in outer bone size or mineral content. Thus, the low strength of elderly male radii arose through different biomechanical mechanisms. Consideration of different strength–age regressions (trajectories) may inform clinical decisions on how best to treat individuals to reduce fracture risk. © 2019 American Society for Bone and Mineral Research.

Despite poor graft integration among some patients that undergo an anterior cruciate ligament (ACL) reconstruction, there has been little consideration of the bone quality into which the ACL femoral tunnel is drilled and the graft is placed. Bone mineral density of the knee decreases following ACL injury. However, trabecular and cortical architecture differences between injured and non‐injured femoral ACL entheses have not been reported. We hypothesize that injured femoral ACL entheses will show significantly less cortical and trabecular mass compared with non‐injured controls. Femoral ACL enthesis explants from 54 female patients (13–25 years) were collected during ACL reconstructive surgery. Control explants (n = 12) were collected from seven donors (18−36 years). Injured (I) femoral explants differed from those of non‐injured (NI) controls with significantly less (p ≤ 0.001) cortical volumetric bone mineral density (vBMD) (NI: 736.1–867.6 mg/cm3; I: 451.2–891.9 mg/cm3), relative bone volume (BV/TV) (NI: 0.674–0.867; I: 0.401–0.792) and porosity (Ct.Po) (NI: 0.133–0.326; I: 0.209–0.600). Injured explants showed significantly less trabecular vBMD (p = 0.013) but not trabecular BV/TV (p = 0.314), thickness (p = 0.412), or separation (p = 0.828). We found significantly less cortical bone within injured femoral entheses compared to NI controls. Lower cortical and trabecular bone mass within patient femoral ACL entheses may help explain poor ACL graft osseointegration outcomes in the young and may be a contributor to the osteolytic phenomenon that often occurs within the graft tunnel following ACL reconstruction.

Objectives: In this proof-of-concept study, we aimed to develop deep-learning-based classifiers to identify rib fractures on frontal chest radiographs in children under two years of age. Methods: This retrospective study included 1311 frontal chest radiographs (radiographs with rib fractures, n = 653) from 1231 unique patients (median age: 4 m). Patients with more than one radiograph were included only in the training set. A binary classification was performed to identify the presence or absence of rib fractures using transfer learning and Resnet-50 and DenseNet-121 architectures. The area under the receiver operating characteristic curve (AUC-ROC) was reported. Gradient-weighted class activation mapping was used to highlight the region most relevant to the deep learning models’ predictions. Results: On the validation set, the ResNet-50 and DenseNet-121 models obtained an AUC-ROC of 0.89 and 0.88, respectively. On the test set, the ResNet-50 model demonstrated an AUC-ROC of 0.84 with a sensitivity of 81% and specificity of 70%. The DenseNet-50 model obtained an AUC of 0.82 with 72% sensitivity and 79% specificity. Conclusions: In this proof-of-concept study, a deep learning-based approach enabled the automatic detection of rib fractures in chest radiographs of young children with performances comparable to pediatric radiologists. Further evaluation of this approach on large multi institutional datasets is needed to assess the generalizability of our results. Advances in knowledge: In this proof-of-concept study, a deep learning-based approach performed well in identifying chest radiographs with rib fractures. These findings provide further impetus to develop deep learning algorithms for identifying rib fractures in children, especially those with suspected physical abuse or non-accidental trauma.

INTRODUCTION Severe traumatic brain injury (TBI) associated with acute subdural hematomas (aSDH) is common and represents around 10% to 20% of all TBI. Predictive models have been used in an attempt to modulate the morbidity and mortality of patient outcomes. We used machine learning (ML) to identify risk factors predictive of in-hospital mortality in the severe TBI patient population with aSDH. METHODS We included 74 patients with severe TBI and aSDH in the analysis. Random forest, ML architecture, was used to create a predictive model of in-hospital mortality with a pre-set precision of 97.4% (RStudio-3.5). A total of 133 input variables including demographics, in-hospital laboratory values, and outcome measures were included and mean accuracy ranks were assessed RESULTS The highest scoring input variables were length of stay, last sodium value collected, last platelet value collected, and Glasgow Coma Scale (GCS) motor exam score on day two. Mean length of stay was significantly shorter in the group of patients that died (4.114 ± 4.241 d vs 22.72 ± 11.72; P < .0001) The mean sodium that was last collected was significantly more elevated in the group of patients who died (139.9 ± 3.299 vs 148.9 ± 8.825 mEq/L; P < .0001). The mean platelet values last collected during the hospitalization were significantly lower in the group of patients who died (440.8 ± 240.4 vs 165.6 ± 113.7 × 109/L; P < .0001). GCS motor exam score at day 2 following the injury was also significantly greater in the survival group (4.872 ± 1.005 vs 2.143 ± 1.574; P < .0001). CONCLUSION ML is an efficient tool that can provide a reasonable level of accuracy in predicting mortality in our patient population. Adequate monitoring of sodium and platelet levels, as well as the GCS motor examination, can promote goal-directed therapy. Integration of ML into the severe TBI algorithm may help early identification of high-risk patients.