基于随机场理论提取DXA图像三维结构特征参数及骨质疏松评估的方法研究

目前骨质疏松评估主要是依据骨密度值，然而大量临床案例表明，仅依靠骨密度值，难以准确预测骨质疏松引起的骨折风险，还需进一步引入重要的骨微结构信息。本项目认为，临床骨密度测量（DXA）得到的X射线二维投影图像蕴含着骨的三维结构信息，因此将对骨三维结构特征参数的提取方法及其在预测骨质疏松骨折风险方面的应用进行理论和实验研究。首先，将利用随机场理论，研究从DXA图像提取骨三维结构参数的方法，重点提取两个参数，它们分别表征微观结构的特征尺寸及其分布特征；其二，研究这两个参数与骨体积比、骨小梁结构之间的关系；其三，采用实验和理论方法，进行累进压缩和疲劳压缩实验，将微观结构和宏观力学测量观察相结合，探究骨三维结构参数对骨的弹性模量、塑性屈服、强度极限等的影响规律，并基于实验结果推导引入特征参数的椎体骨本构关系；其四，与本科室临床数据比较，初步验证该方法的有效性，为提高临床骨折风险预测的准确性提供新思路。

Dual-energy X-ray absorptiometry (DXA) is the most widely used clinical mean to diagnose osteoporosis. However, both clinical pratice and research showed that Bone mineral density (BMD) is not sufficient to predict bone fracture risk. Bone microstructure is an essential factor to be considered during diagnosis of osteoporois. This study will develop the method to extract the three-dimensional structural parameters and improve the accuracy to predict osteoporotic fracture. First, we will start with obtaining characteristic parameters of vertebra using the stochastic theory. Second, we will investigate the relation between the parameters and bone structures, such as bone volume fraction, trabecular thickness, number and spacing. Third, we will conduct compressive progressive loading and fatigue loading, and reveal how the change of bone structure influences the elastic modulus, strength, plastic deformation, and constitutive relation of vertebra. Finally,we will apply the technique to compare with the clincial data available in our hospital to test its efficacy and provide a new way to better predict osteoporotic fracture.