多尺度增强算法在肺结节计算机辅助检测中的应用探讨  

The application of multi-scale enhancement algorithm in lung nodule computer-aided detection

作　　者：张娟[1] 侯庆锋[1] 王昌元[2] 鲁雯[1] 张会如[1] 

机构地区：[1]泰山医学院放射学院,山东泰安271016 [2]泰山医学院信息工程学院,山东泰安271016

出　　处：《泰山医学院学报》2009年第6期410-412,共3页Journal of Taishan Medical College

摘　　要：目的研究多尺度增强算法对肺小结节的增强能力及将其作为肺结节计算机辅助检测的处理方法的可行性。方法针对肺结节的形态特点,采用高斯函数模拟肺结节,运用多尺度图像增强滤波器算法,增强肺结节提取兴趣区,供后续的分类判别使用。结果通过对肺部CT图像的应用,说明多尺度增强算法能很好地检测疑似肺结节兴趣区。结论多尺度增强算法对帮助医师检测肺结节有明显的作用,是一种有效的图像预处理方法,对肺结节的计算机辅助检测有较大应用价值。Objective： To study multi-scale enhancement algorithm and its application in lung nodule computer-aided detection as a preproeessing method. Methods ： Because most lung nodules were circle or like circle and their density distributions were Ganssian distribution, lung nodule was simulated and our muhi-scale enhancement filter was constructed by Gaussian to enhance nodules with varies diameters. Results： The application results indicated that our muhi-scale enhancement filter effectively enhanced nodules with varies diameters. Conclusion： Muhi-scale enhancement algorithm is an effective enhancement filter for lung nodule detection.

关 键 词：多尺度增强 计算机辅助检测 高斯函数 感兴趣区 

分 类 号：R816.4[医药卫生—放射医学]