Risk Assessment of Hip Fracture Based on Machine Learning

Identifying patients with high risk of hip fracture is a great challenge in osteoporosis clinical assessment. Bone Mineral Density (BMD) measured by Dual-Energy X-Ray Absorptiometry (DXA) is the current gold standard in osteoporosis clinical assessment. However, its classification accuracy is only around 65%. In order to improve this accuracy, this paper proposes the use of Machine Learning (ML) models trained with data from a biomechanical model that simulates a sideways-fall. Machine Learning (ML) models are models able to learn and to make predictions from data. During a training process, ML models learn a function that maps inputs and outputs without previous knowledge of the problem. The main advantage of ML models is that once the mapping function is constructed, they can make predictions for complex biomechanical behaviours in real time. However, despite the increasing popularity of Machine Learning (ML) models and their wide application to many fields of medicine, their use as hip fracture predictors is still limited. This paper proposes the use of ML models to assess and predict hip fracture risk. Clinical, geometric, and biomechanical variables from the finite element simulation of a side fall are used as independent variables to train the models. Among the different tested models, Random Forest stands out, showing its capability to outperform BMD-DXA, achieving an accuracy over 87%, with specificity over 92% and sensitivity over 83%.

Periprosthetic femoral fractures in sideways fall configuration: comparative numerical analysis of the influence of femoral stem design

Introduction: The aim of this study was to investigate the mechanisms of periprosthetic fractures occurring as a result of a sideways fall in total hip arthroplasty patients, and to compare the predictions of numerical models in terms of load distribution on the implanted femur with clinical data. Materials and methods: 3 numerical models were built: 1 for intact femur and 2 for implanted femur with a straight stem (resembling PBF, Permedica) and with an anatomical stem (resembling ABG II, Stryker). 4 loading configurations were simulated; 1 simulates a vertical load, and 3 simulate a fall with impact on the greater trochanter in different directions. Stress state calculated in the implanted femur was compared for the 2 models with reference to the intact case. These were compared with clinical data collected at a single centre (Istituto Ortopedico Gaetano Pini, Milan, Italy) where 41 patients were investigated after periprosthetic fracture: 26 patients had a straight uncemented stem and 15 an anatomical uncemented stem. Results: The maximum calculated strain in compression in the case of ABG II implanted femur was 2 times higher than in the presence of PBF stem in the vertical loading configuration. For configurations of sideways fall, in both models, there was a progressive increase of stress state in the bone with increasing angle. Simulations of sideways fall elicited results in accordance with clinical observations: due to the peculiar stem design and consequent state of stress in the bone, anatomical stems seem to induce trochanteric fractures more frequently, while for straight stems type B fractures are more likely to occur. Conclusions: Clinical findings confirmed numerical model predictions: stem design seems to highly influence distribution of stress in the bone and consequent localisation of the fracture site.

Body Anthropometry and Bone Strength Conjointly Determine the Risk of Hip Fracture in a Sideways Fall

We hypothesize that variations of body anthropometry, conjointly with the bone strength, determine the risk of hip fracture. To test the hypothesis, we compared, in a simulated sideways fall, the hip impact energy to the energy needed to fracture the femur. Ten femurs from elderly donors were tested using a novel drop-tower protocol for replicating the hip fracture dynamics during a fall on the side. The impact energy was varied for each femur according to the donor’s body weight, height and soft-tissue thickness, by adjusting the drop height and mass. The fracture pattern, force, energy, strain in the superior femoral neck, bone morphology and microarchitecture were evaluated. Fracture patterns were consistent with clinically relevant hip fractures, and the superior neck strains and timings were comparable with the literature. The hip impact energy (11 – 95 J) and the fracture energy (11 – 39 J) ranges overlapped and showed comparable variance (CV = 69 and 61%, respectively). The aBMD-based definition of osteoporosis correctly classified 7 (70%) fracture/non-fracture cases. The incorrectly classified cases presented large impact energy variations, morphology variations and large subcortical voids as seen in microcomputed tomography. In conclusion, the risk of osteoporotic hip fracture in a sideways fall depends on both body anthropometry and bone strength.