基于3D-CT图像的髋脱位定量智能诊断方法

根据正常状态下股骨头与髋臼月状面所在球面同心的原理，结合三维图像分割与球面拟合两种主要思想以及细节的算法处理，提出一种有效的髋脱位定量辅助诊断方法.首先将髋臼与股骨头三维分割，获取各自的独立数据；提取髋臼与股骨头表面数据并剔除误差点后，利用最小二乘法分别拟合出股骨头表面和髋臼月状面所在的球面，并通过两球面的球心距数值确定髋臼是否脱位.实验结果表明，此方法判断结果有效，同时可计算出髋臼复位需要移动的距离，提高诊断精度.

Quantitative and Intelligent Diagnosis Method of Hip Dislocation Based on 3D-CT Images

Considering that the spherical surface of the femoral head should be approximately concentric with that of the acetabulum， an effective method based on three dimensional image segmentation and spherical surface fitting was proposed， with some detailed algorithm designed. First， the three dimensional segmentation was implemented on the femur and acetabulum to obtain respective independent data. After extracting the valid data of the femoral head and acetabulum boundary and removing the error points， the least square method was applied to fit the spherical surface of the femoral head and the surface of the acetabulum. Then the respective sphere center distance between two spherical surfaces was calculated， thus determining whether the acetabulum is dislocated or not. The results of experiments show that the method proposed can provide the distance for hip replacement and improve the diagnostic accuracy.