青光眼是不可逆性视力损害的主要原因, 对前房角进行分类可以方便医生诊断。提出了基于Spaeth分级法的房角分类算法。在超声生物显微镜图像中, 利用边界搜索的方法确定房角粗定位点, 并在灰度图像中完成房角细定位。计算出分类所需的各项参数, 最后基于Spaeth分级法将前房角分为开、闭、窄三类。所提方法与现有方法相比具有两个主要特征: 第一, 自动完成房角顶点的定位, 且对于不同类型的房角结构都有效。第二, 解决房角多分类问题, 分类依据与临床分级方法相似。实验表明, 方法具有较高的准确性和效率, 具有较大的实用性和应用参考价值。

Glaucoma is the main cause of irreversible visual impairment and can be detected or diagnosed by classifying anterior chamber angle (ACA). An automatic chamber angle classification algorithm based on Spaeth classification is proposed. In ultrasound biomicroscopy (UBM) imaging, the critical position of the binary image is located according to the boundary search method, and the vertice of the ACA is precisely located in the grayscale image.This method calculates the parameters required for classification. Finally, based on Spaeth classification, the ACA is divided into three categories: open, closed and narrow. This method has two principal features compared with existing methods. First, the vertex positioning is completely automatic and effective in different ACA structures. Second, the multi-classification problem is completed, and the classification basis is similar to clinical judgment. Experiments show the method has high accuracy and efficiency and has great practicability and application reference value.