桡骨远端骨折夹板外固定的骨折缝应力预测

目的 运用静力学分析、试验设计法以及布谷鸟-支持向量机预测建立关于 C1_1、C2_1、C3_1 型桡骨远端骨折夹板固定不同厚度压垫下骨折缝应力关系的数学模型。 方法 在 CT 影像中获得前臂中各个组织的形状,建立对应的实体模型,进行降噪、装配等过程实现建模;在 Workbench 中根据实验测定的载荷数据对模型施加载荷,获得一种情况下的骨组织力学情况;再通过试验设计法对多种施加载荷情况下的骨组织力学情况进行模拟;最后,通过布谷鸟改进支持向量回归预测,获得关于各骨折类型的数学模型。 结果 对 B 曲面划分的 12 个压垫固定位置施加载荷,通过试验设计法对 460 个样本在 Workbench 中 DOE 模块中进行求解,460 个样本中随机抽取 400 个作为训练集得到数学模型,用剩余 60 个样本作为测试集与数学模型预测结果进行对比,即可建立高精度的有关夹板固定中压垫固定位置上不同载荷和骨折缝处载荷情况的数学模型。 通过已知压垫固定区域施加不同厚度压垫的固定力对骨折缝应力的预测,预测值与仿真值进行对比模型预测精度最高达到 99. 66% ,最低达到 95. 653% 。 从精度上分析,压垫载荷与骨折缝处载荷存在模糊的数学关系,即支持向量机预测模型有效。 结论 通过研究压垫固定作用方式和压垫固定力对骨折缝的影响,有利于传统固定夹板与现代科学技术相结合,为临床骨折治疗桡骨远端骨折时固定器的选择和外固定器械设计提供参考。

Prediction of Fracture Suture Stress for Distal Radius Fractures Fixed by the Splint

Objective To establish a mathematical model for fracture suture load relationship of C1 _1, C2 _1, C3_1 type distal radius fractures fixed by the splint with different pad thicknesses by using static analysis, designof experiment (DOE) and cuckoo-support vector machine prediction. Methods The shape of each tissue in the forearm was obtained from CT images, the corresponding solid models were established, and the modeling wasachieved by noise reduction, assembly processes. The mechanical situation of bone tissues under one conditionwas obtained by applying load to the model according to load data measured by the Workbench. Then throughthe DOE method, the mechanical conditions of bone tissues under various conditions of the applied load weresimulated. Finally, the mathematical models for various fracture types were obtained by cuckoo search-support vector regression (CS-SVR) prediction. Results The load was applied to the fixed position of 12 pressure padsdivided by the B surface, and 460 samples were solved in DOE module in Workbench by the DOE method, and400 of the 460 samples were randomly selected as the training set to obtain the mathematical model. Theremaining 60 samples were used as the test set for comparison with prediction result of the mathematical model.The high precision mathematical models for different loads on the fixed position of the splint and the load atfracture suture could be established. The highest prediction accuracy of the model was 99. 659% , and the lowestwas 95. 653% . From the accuracy analysis, there was a fuzzy mathematical relationship between the pad loadand fracture suture load, indicating that the support vector machine prediction model was effective. ConclusionsThe study on the effect of pressure splint fixation mode and pressure splint stress on fracture suture is beneficialfor the combination of traditional fixation splint and modern science and technology, and it can provide referencesfor the selection of fixators and the design of external fixation devices in clinical treatment of distal radiusfractures.