基于径向梯度指数的肺结节假阳性筛除方法

False Positive Reduction Method for Lung Nodules Based on Radial Gradient Index

针对肺结节自动检测中假阳性率高影响检测结果的问题,提出一种筛除方法。采用基于径向梯度指数的方法降低假阳性率,生成CT图像的基于径向梯度指数图,设定基于径向梯度指数滤波器滤除假阳性结节。实验结果表明,该方法能有效降低肺结节自动检测的假阳性率。

Aiming at CT lung nodule detection,the false positive rate is a major disturb factor.This paper proposes a Radial Gradient Index（RGI）-based method.Based on RGI map of CT image,RGI filter is set to reduce false-positive nodules.Experimental results show this method can efficiently reduce the False Positive Rate（FPR） in automated lung nodule detection.