提出一种基于Gabor特征的非监督、全自动的椎间盘定位与退行性变分级算法。首先通过对一系列脊柱Gabor特征图像的处理,得到脊柱和椎间盘的边缘信息,并基于脊柱边缘信息提取脊柱区域;然后根据以上信息及椎间盘位置的先验信息,在脊柱区域定位椎间盘;最后根据定位结果和Gabor系数图,结合椭圆拟合的方法,得到髓核和纤维环灰度信息,并根据灰度和几何信息实现分级。通过37个病人MRI数据,验证算法的准确率和可行性,其中定位方法的准确率达96.6%,在与已有方法准确率相当的情况下降低了算法复杂度,并将定位精度提高至1.46 mm,而分级算法可以实现前5级的退行性变评定。
An unsupervised method based on Gabor feature for localizing intervertebral discs( IVDs)automatically and classifying disc degeneration was proposed in this paper. At first,a series of Gabor-filtered spine images were obtained in order to extract the information of spine edges and disc edges based on which the areas of spines could be determined. After that,on the basis of the prior knowledge,the exact location of IVDs in the spines were calculated. Finally,associating localization results with Gabor coefficients,the classification of IVD degeneration was realized based on gray information of nucleus pulposus and annulus fibrosis which were obtained by ellipse fitting. Experiments were performed in a dataset of 37 patients,the results showed that our method was simpler than the existing methods while with a similar accuracy of 96. 6%,and it increased the precision to 1. 46 mm. Besides,the degeneration classification of level 1- 5 could be realized by our classification algorithm.