摘要
目的建立一种基于卷积神经网络的脑血肿分割算法,探讨算法与手动分割结果的一致性。方法纳入中国颅内出血影像数据库中146例头部CT平扫影像图片,采用随机数字表法分为训练集(90例)、测试集(26例)和验证集(30例),验证集采用手动分割、算法分割、精确多田公式和传统多田公式共4种方法对血肿体积进行测量,以手动分割为"金标准",分别对其他3种算法进行一致性检验。结果与多田公式方法相比,算法分割的百分误差最小,为15.54(8.41,23.18)%,组内相关系数最高,为0.983;Bland-Altman一致性检测显示,93.33%的数据在95%一致性界限(95%LoA),且其95%LoA最窄,为-6.46~5.97 ml。算法分割的百分误差在不同血肿形态、体积比较中差异无统计学意义(均P>0.05)。结论卷积神经网络构建的算法分割具有一定的临床应用前景,但仍需更大样本的临床试验加以验证。
Objective To establish an automatic segmentation algorithm using convolution neural network,and to validate the consistency between the algorithm and manual segmentation.Methods One hundred and forty-six CT scans of intracerebral hemorrhage(ICH)were included from Chinese Intracranial Hemorrhage Image Database(CICHID).They were randomly divided into training set(n=90),testing set(n=26)and validation set(n=30).All CT scans were manual segmentation.Training set and testing set were used for algorithm training.The validation set was measured by four methods including manual segmentation,algorithm segmentation,accurate Tada formula and traditional Tada formula.The consistency test was performed.Results Compared with the Tada formula methods,the percentage error of algorithm values was the smallest 15.54(8.41,23.18)%,and algorithm agreement with the manual reference was the strongest(correlation coefficient 0.983).Bland-Altman analysis showed that 93.33%of the data was within the 95%limits of agreement(95%LoA),and 95%LoA was narrow(-6.46-5.97 ml).No significant differences were found in size and shape(P>0.05,for all).Conclusions The algorithm using convolutional neural network has a certain application prospect,but it needs still more validation in large sample research.
作者
常健博
姜燊种
陈显金
骆嘉希
李沃霖
张庆华
魏俊吉
石林
冯铭
王任直
CHANG Jian⁃bo;JIANG Shen⁃zhong;CHEN Xian⁃jin;LOK Ka⁃hei;LEE Yuk⁃lam;ZHANG Qing⁃hua;WEI Jun⁃ji;SHI Lin;FENG Ming;WANG Ren⁃zhi(Department of Neurosurgery,Peking Union Medical College Hospital,Chinese Academy of Medical Sciences and Peking Union Medical College,Beijing 100730,China;Department of Neurosurgery,Union Shenzhen Hospital(Nanshan Hospital),Huazhong University of Science and Technology,Shenzhen 518051,Guangdong,China;Shenzhen BrainNow Research Institute,Shenzhen 518000,Guangdong,China;Department of Imaging and Interventional Radiology,the Chinese University of Hong Kong,Hongkong 999077,China)
出处
《中国现代神经疾病杂志》
CAS
北大核心
2020年第7期585-590,共6页
Chinese Journal of Contemporary Neurology and Neurosurgery
基金
北京市自然科学基金资助项目(项目编号:7182137)
中国医学科学院医学与健康科技创新工程重大协同创新项目(项目编号:2017-I2M-3-014)
中国医学科学院北京协和医学院研究生教育教学立项项目(项目编号:10023201900107)
关键词
脑出血
人工智能
神经网络(计算机)
体层摄影术
X线计算机
Cerebral hemorrhage
Artificial intelligence
Neural networks(computer)
Tomography,X-ray computed
