Introduction. The increased prevalence of osteoporosis in the patient population undergoing reverse shoulder arthroplasty (RSA) results in significantly increased complication rates. Mainly demographic and clinical predictors are currently taken into the preoperative assessment for risk stratification without quantification of preoperative computed tomography (CT) data (e.g. bone density). It was hypothesized that preoperative CT bone density measures would provide objective quantification with subsequent classification of the patients’ humeral bone quality. Methods. Thirteen bone density parameters from 345 preoperative CT scans of a clinical RSA cohort represented the data set in this study. The data set was divided into testing (30%) and training data (70%), latter included an 8-fold cross validation. Variable selection was performed by choosing the variables with the highest descriptive value for each correlation clustered variables. Machine learning models were used to improve the clustering (Hierarchical Ward) and classification (Support Vector Machine (SVM)) of bone densities at risk for complications and were compared to a conventional statistical model (Logistic Regression (LR)). Results. Clustering partitioned this cohort (training data set) into a high bone density subgroup consisting of 96 patients and a low bone density subgroup consisting of 146 patients. The optimal number of clusters (n = 2) was determined based on optimization metrics. Discrimination of the cross validated classification model showed comparable performance for the training (accuracy=91.2%; AUC=0.967) and testing data (accuracy=90.5 %; AUC=0.958) while outperforming the conventional statistical model (Logistic Regression (LR)). Local interpretable model-agnostic explanations (LIME) were created for each patient to explain how the predicted output was achieved. Conclusion. The trained and tested model provides preoperative information for surgeons treating patients with potentially poor bone quality. The use of machine learning and patient-specific calibration showed that multiple 3D bone density scores improved accuracy for objective preoperative bone quality assessment

The objective of this study was to use patient-specific finite element modeling to measure the 3D interfragmentary strain environment in clinically realistic fractures. The hypothesis was that in the early post-operative period, the tissues in and around the fracture gap can tolerate a state of strain in excess of 10%, the classical limit proposed in the Perren strain theory. Eight patients (6 males, 2 females; ages 22–95 years) with distal femur fractures (OTA/AO 33-A/B/C) treated in a Level I trauma center were retrospectively identified. All were treated with lateral bridge plating. Preoperative computed tomography scans and post-operative X-rays were used to create the reduced fracture models. Patient-specific materials properties and loading conditions (20%, 60%, and 100% body weight (BW)) were applied following our published method.[1]. Elements with von Mises strains >10% are shown in the 100% BW loading condition. For all three loading scenarios, as the bridge span increased, so did the maximum von Mises strain within the strain visualization region. The average gap closing (Perren) strain (mean ± SD) for all patient-specific models at each body weight (20%, 60%, and 100%) was 8.6% ± 3.9%, 25.8% ± 33.9%, and 39.3% ± 33.9%, while the corresponding max von Mises strains were 42.0% ± 29%, 110.7% ± 32.7%, and 168.4% ± 31.9%. Strains in and around the fracture gap stayed in the 2–10% range only for the lowest load application level (20% BW). Moderate loading of 60% BW and above caused gap strains that far exceeded the upper limit of the classical strain rule (<10% strain for bone healing). Since all of the included patients achieved successful unions, these findings suggest that healing of distal femur fractures may be robust to localized strains greater than 10%