生物医学工程研究2016，Vol.35Issue(4)：290-293,4.DOI:10.19529/j.cnki.1672-6278.2016.04.15
基于B P神经网络的脑肿瘤MRI图像分割
Brain Tumor Segmentation of MRI based on BP Neural Network
摘要
Abstract
To distinguish normal tissue area and the tumor area in MRI brain images by using BP neural network technology First-ly,the sample image segmentation with the image processing software was made.Secondly,the sample image segmentation was put into the neural network model as the known sample for training.Finally,the unknown MRI with the trained neural network was classified. Applying the BP neural network in the segmentation of brain MRI could identify the difference between the tumor and the surrounding normal tissues effectively.To improve the accuracy,we input samples of fuzzy area with specific training in the model and used special filtering method.The BP neural network can realize accurate and effective segmentation of brain MRI,but it’s important to choose in-put sample correctly and use special filtering method.关键词
BP神经网络/MRI图像/脑肿瘤/滤波/图像分割Key words
BP neural network/MRI image/Brain tumor/Filter/Image segmentation分类
医药卫生引用本文复制引用
王锦程,郁芸,杨坤,胡新华..基于B P神经网络的脑肿瘤MRI图像分割[J].生物医学工程研究,2016,35(4):290-293,4.基金项目
南京市医学科技发展项目（YKK12125、201303009）；江苏省大学生实践创新训练计划基金资助项目（201410312057X）；南京医科大学“十二五”教育研究课题（JYY2015081）。 （）
