帮助 关于我们

返回检索结果

基于多中心MRI的3D-ResNet101深度学习模型预测脑胶质瘤术前分级的研究
The MRI-based 3D-ResNet101 deep learning model for predicting preoperative grading of gliomas: A multicenter study

查看参考文献30篇

李大瑞 1,2,3   胡万均 1,2   刘光耀 1,2   甘铁军 1,2   马来阳 1,2,3   张静 1,2 *  
文摘 目的术前准确无创预测胶质瘤分级仍然具有挑战性。基于常规T2WI图像开发一种鲁棒性强的残差神经网络(Residual Networks, ResNet)深度学习模型以预测脑胶质瘤术前病理分级。材料与方法回顾性分析919例经病理证实为胶质瘤患者的术前T2WI图像,其中708例为2014年6月至2021年4月在兰州大学第二医院收治的患者数据,211例来源于癌症影像档案(The Cancer Imaging Archive, TCIA)数据库。TCIA数据集又被细分为开发集(n=135)和独立测试集(n=76),将兰州大学第二医院数据集和TCIA开发集的数据按7∶3随机分为训练集(n=590)和测试集(n=253),基于T2WI图像构建3D-ResNet101深度学习模型。训练后的模型在测试集和独立测试集进行验证,并通过宏观F1分数、准确率(accaruy, ACC)及受试者工作特征(receiver operating characteristic, ROC)曲线对模型效能进行评估。结果基于T2WI构建的3D-ResNet101深度学习模型在训练集和测试集ACC分别为99%、95%,F1分数分别为99%、95%,ROC曲线下面积(area under the curve, AUC)分别为0.98、0.97;独立测试集ACC为83%、F1分数为83%、AUC为0.89。结论基于T2WI图像的3D-ResNet101深度学习模型预测高、低级别胶质瘤具有较高的准确性、鲁棒性。该方法可用于术前胶质瘤分级的无创预测,并有助于提升患者临床管理的有效性。
其他语种文摘 Objective: The preoperative accurate and non-invasive prediction of glioma grading remains challenging. Developing a robust Residual Networks (ResNet) deep learning model based on conventional T2WI images to predict preoperative pathological grading of gliomas. Materials and Methods: A retrospective analysis of preoperative T2WI images of 919 patients with pathologically confirmed glioma, of which 708 were data from patients enrolled at the Second Hospital of Lanzhou University from June 2014 to April 2021 and 211 were derived from The Cancer Imaging Archive (TCIA) database. The TCIA dataset was subdivided into a development set (n=135) and an independent test set (n=76). The data from the Second Hospital of Lanzhou University dataset and the TCIA development set were randomly split 7∶3 into a training set (n=590) and a test set (n=253) to construct a 3D-ResNet101 deep learning model based on T2WI images. After the training, the models were validated on the test set and independent test set, where the model efficacy was assessed by macro F1 scores, accuracy (ACC), and receiver operating characteristic (ROC) curves. Results: The 3D-ResNet101 deep learning model constructed based on T2WI had ACCs of 99% and 95% in the training and test sets, respectively; The F1 scores were 99% and 95%, respectively; the area under the ROC curve (AUC) were 0.98 and 0.97, respectively; the ACC of the independent test set was 83%, the F1 score was 83%, and the AUC was 0.89. Conclusions: The 3D-ResNet101 deep learning model based on T2WI images predicts high- and low-grade gliomas with high accuracy and robustness. The method can be used for the non-invasive prediction of preoperative glioma grading as well as helping to improve the effectiveness of clinical management of patients.
来源 磁共振成像 ,2023,14(5):25-30 【核心库】
DOI 10.12015/issn.1674-8034.2023.05.006
关键词 胶质瘤 ; 3D-残差神经网络 ; 深度学习 ; 磁共振成像 ; T2加权成像
地址

1. 兰州大学第二医院核磁共振科, 兰州, 730030  

2. 甘肃省功能及分子影像临床医学研究中心, 甘肃省功能及分子影像临床医学研究中心, 兰州, 730030  

3. 兰州大学第二临床医学院, 兰州, 730030

语种 中文
文献类型 研究性论文
ISSN 1674-8034
学科 临床医学;肿瘤学
基金 甘肃省卫生健康行业科研项目 ;  兰州大学第二医院"萃英科技创新"项目
文献收藏号 CSCD:7476440

参考文献 共 30 共2页

1.  Jiang T. Clinical practice guidelines for the management of adult diffuse gliomas. Cancer Lett,2021,499:60-72 CSCD被引 50 次    
2.  Louis D N. The 2016 World Health Organization Classification of Tumors of the Central Nervous System: a summary. Acta Neuropathol,2016,131(6):803-820 CSCD被引 513 次    
3.  Su C Q. Intergrating conventional MRI, texture analysis of dynamic contrast-enhanced MRI, and susceptibility weighted imaging for glioma grading. Acta Radiol,2019,60(6):777-787 CSCD被引 7 次    
4.  张斌. 深度学习在脑胶质瘤影像学的研究进展. 中国医学物理学杂志,2021,38(8):1048-1052 CSCD被引 2 次    
5.  Shin I. Development and Validation of a Deep Learning-Based Model to Distinguish Glioblastoma from Solitary Brain Metastasis Using Conventional MR Images. AJNR Am J Neuroradiol,2021,42(5):838-844 CSCD被引 5 次    
6.  Ahammed Muneer K V. Glioma Tumor Grade Identification Using Artificial Intelligent Techniques. J Med Syst,2019,43(5):113 CSCD被引 4 次    
7.  He K. Deep residual learning for image recognition. Proceedings of IEEE Conference on Computer Vision and Pattern Recognition,2016:770-778 CSCD被引 94 次    
8.  Korfiatis P. Residual Deep Convolutional Neural Network Predicts MGMT Methylation Status. J Digit Imaging,2017,30(5):622-628 CSCD被引 11 次    
9.  Clark K. The Cancer Imaging Archive (TCIA): maintaining and operating a public information repository. J Digit Imaging,2013,26(6):1045-1057 CSCD被引 39 次    
10.  Choi Y S. Fully automated hybrid approach to predict the IDH mutation status of gliomas via deep learning and radiomics. Neuro Oncol,2021,23(2):304-313 CSCD被引 17 次    
11.  Zha C. Neutrophil extracellular traps mediate the crosstalk between glioma progression and the tumor microenvironment via the HMGB1/RAGE/IL-8 axis. Cancer Biol Med,2020,17(1):154-168 CSCD被引 18 次    
12.  Mzoughi H. Deep Multi-Scale 3D Convolutional Neural Network (CNN) for MRI Gliomas Brain Tumor Classification. J Digit Imaging,2020,33(4):903-915 CSCD被引 7 次    
13.  Singh G. Radiomics and radiogenomics in gliomas: a contemporary update. Br J Cancer,2021,125(5):641-657 CSCD被引 8 次    
14.  Zhuge Y. Automated glioma grading on conventional MRI images using deep convolutional neural networks. Med Phys,2020,47(7):3044-3053 CSCD被引 9 次    
15.  Yang Y. Glioma Grading on Conventional MR Images: A Deep Learning Study With Transfer Learning. Front Neurosci,2018,12:804 CSCD被引 8 次    
16.  Li Y. Molecular subtyping of diffuse gliomas using magnetic resonance imaging: comparison and correlation between radiomics and deep learning. Eur Radiol,2022,32(2):747-758 CSCD被引 4 次    
17.  Khawaldeh S. Noninvasive grading of glioma tumor using magnetic resonance imaging with convolutional neural networks. Applied Sciences,2017,8(1):27 CSCD被引 2 次    
18.  Artzi M. Automatic segmentation, classification, and follow-up of optic pathway gliomas using deep learning and fuzzy c-means clustering based on MRI. Med Phys,2020,47(11):5693-5701 CSCD被引 2 次    
19.  Bangalore Yogananda C G. A novel fully automated MRI-based deep-learning method for classification of IDH mutation status in brain gliomas. Neuro Oncol,2020,22(3):402-411 CSCD被引 8 次    
20.  Yu J. 2D CNN versus 3D CNN for false-positive reduction in lung cancer screening. J Med Imaging (Bellingham),2020,7(5):051202 CSCD被引 3 次    
引证文献 2 篇

1 陈晓华 基于多序列MRI的卷积神经网络预测胶质瘤MGMT启动子甲基化状态 磁共振成像,2023,14(8):34-39,78
CSCD被引 1 次

2 张婷婷 基于增强层次对称点图像分析和深度残差网络的水电机组故障诊断 水利学报,2023,54(11):1380-1391
CSCD被引 2 次

显示所有2篇文献

论文科学数据集
PlumX Metrics
相关文献

 作者相关
 关键词相关
 参考文献相关

版权所有 ©2008 中国科学院文献情报中心 制作维护：中国科学院文献情报中心
地址：北京中关村北四环西路33号 邮政编码：100190 联系电话：(010)82627496 E-mail:cscd@mail.las.ac.cn 京ICP备05002861号-4 | 京公网安备11010802043238号