了解肺裂解剖结构特征在定位和评估肺部疾病方面具有非常重要的作用。在计算机断层扫描图像中肺裂经常会受到形变、体素效应和噪声等影响,使得肺裂检测难度很大。为了解决此问题,提出一种基于形状特征的肺裂检测算法。首先融合肺裂幅度信息和方向信息在增强肺裂的同时高效率的抑制噪声;然后利用区域属性分析算法去除气管、血管等噪声来识别肺裂;最后采用表面曲率算法去除黏连噪声,达到分割肺裂的目的。该算法在公开数据集LOLA11上进行了验证。与人工参考对比,本文算法分割的肺裂的F₁-score中值为0.8451。实验结果表明,本文算法能够高效率的分割肺裂。

Knowledge of pulmonary fissure anatomy plays an important role in localization of lesions and evaluation of lung disease. In computed tomography images, pulmonary fissure detection is an intricate task due to factors such as pathological deformation, partial volume effect and noise. To solve the problem, a novel method based on shape features is proposed for pulmonary fissure detection. Firstly, the orientation information and magnitude information of pulmonary fissures are fused to enhance pulmonary fissures and suppress interferences. Then region property analysis algorithm is used to remove interferences like airways and vessels for pulmonary fissure identification. Finally, surface curvature approach is utilized to remove adhering interferences for pulmonary fissure segmentation. The performance of the proposed method is validated in experiments with a publicly available LOLA11 dataset. Compared with manual references, the proposed method acquired a high median F1-score of 0.8451. Experimental results show that the proposed method has a good performance in pulmonary fissure segmentation.