Machine Vision and Medical Image Processing (MVMIP) Lab., Department of Biomedical Engineering, Faculty of Electrical Engineering, K.N. Toosi University of Technology, Tehran, Iran.
Machine Vision and Medical Image Processing (MVMIP) Lab., Department of Biomedical Engineering, K.N. Toosi University of Technology, P.O. Box 163151355, Tehran, Iran.
J Digit Imaging. 2018 Oct;31(5):727-737. doi: 10.1007/s10278-018-0076-9.
Airway and vessel characterization of bronchiectasis patterns in lung high-resolution computed tomography (HRCT) images of cystic fibrosis (CF) patients is very important to compute the score of disease severity. We propose a hybrid and evolutionary optimized threshold and model-based method for characterization of airway and vessel in lung HRCT images of CF patients. First, the initial model of airway and vessel is obtained using the enhanced threshold-based method. Then, the model is fitted to the actual image by optimizing its parameters using particle swarm optimization (PSO) evolutionary algorithm. The experimental results demonstrated the outperformance of the proposed method over its counterpart in R-squared, mean and variance of error, and run time. Moreover, the proposed method outperformed its counterpart for airway inner diameter/vessel diameter (AID/VD) and airway wall thickness/vessel diameter (AWT/VD) biomarkers in R-squared and slope of regression analysis.
支气管扩张症在囊性纤维化（CF）患者肺部高分辨率计算机断层扫描（HRCT）图像中的气道和血管特征对于计算疾病严重程度的评分非常重要。我们提出了一种混合和进化优化的阈值和基于模型的方法，用于 CF 患者肺部 HRCT 图像中气道和血管的特征描述。首先，使用增强的基于阈值的方法获得气道和血管的初始模型。然后，通过使用粒子群优化（PSO）进化算法优化其参数来拟合实际图像。实验结果表明，与对照组相比，该方法在 R 平方、误差均值和方差以及运行时间方面表现出色。此外，在 R 平方和回归分析的斜率方面，该方法在气道内直径/血管直径（AID/VD）和气道壁厚度/血管直径（AWT/VD）生物标志物方面也优于对照组。
