您的位置：山东大学 -> 科技期刊社 -> 《山东大学学报（理学版）》

《山东大学学报(理学版)》 ›› 2025, Vol. 60 ›› Issue (7): 13-21.doi: 10.6040/j.issn.1671-9352.0.2023.550

• • 上一篇    

基于深度真值发现的胶质瘤基因状态预测方法

赵钰琳,梁峰宁,赵藤,曹亚茹,王淋,朱红*   

  1. 徐州医科大学医学信息与工程学院, 江苏 徐州 221004
  • 发布日期:2025-07-01
  • 通讯作者: 朱红(1970— ),女,教授,博士,研究方向为人工智能、机器学习、深度学习、模式识别、智能医学图像处理、智能医学信息处理. ;E-mail:zhuhong@xzhmu.edu.cn
  • 作者简介:赵钰琳(1997— ),女,硕士研究生,研究方向为智能医学图像处理. E-mail:301109110887@stu.xzhmu.edu.cn*通信作者:朱红(1970— ),女,教授,博士,研究方向为人工智能、机器学习、深度学习、模式识别、智能医学图像处理、智能医学信息处理. E-mail:zhuhong@xzhmu.edu.cn
  • 基金资助:
    国家自然科学基金资助项目(62102345);江苏省卫生健康委医学科研项目(Z2020032);徐州市重点研发计划资助项目(KC22117)

Method for glioma gene status prediction based on deep truth discovery

ZHAO Yulin, LIANG Fengning, ZHAO Teng, CAO Yaru, WANG Lin, ZHU Hong*   

  1. School of Medical Information and Engineering, Xuzhou Medical University, Xuzhou 221004, Jiangsu, China
  • Published:2025-07-01

PDF (PC)

2

摘要： 针对目前基于磁共振成像(magnetic resonance imaging, MRI)的胶质瘤异柠檬酸脱氢酶1(isocitrate dehydrogenase 1, IDH1)基因状态预测的深度学习模型中存在的深度网络特征提取不全面、模型存在固有的不确定性等问题,提出基于改进的残差网络(residual network, ResNet)与真值发现的TDA-ResNet(truth discovery divided attention-ResNet)模型。通过分散注意力机制优化ResNet网络模型架构,提取胶质瘤影像局部与全局特征,对胶质瘤IDH1基因状态进行预测;同时在模型中融入真值发现算法,对作为预测结果的深度特征向量进行不确定性校准,提高模型预测准确率。实验数据收集自徐州医科大学附属医院部分胶质瘤患者的MR影像及癌症影像档案(the cancer imaging archive, TCIA)公有数据集。TDA-ResNet模型在徐州医科大学附属医院胶质瘤MR影像数据集及TCIA数据集中的实验准确率分别为95.73%和94.3%。实验结果表明,TDA-ResNet模型可实现对脑胶质瘤IDH1基因状态无创预测及不确定性校准,其性能优于现有的IDH1基因状态深度学习预测模型,对脑胶质瘤临床诊疗具有重要意义。

关键词: 脑胶质瘤, 异柠檬酸脱氢酶1, 深度学习, 真值发现, 不确定性校准

Abstract: Aiming at the problems of incomplete deep network feature extraction and inherent uncertainty of the model in the current deep learning model for the prediction of glioma isocitrate dehydrogenase 1(IDH1)gene status based on Magnetic Resonance Imaging(MRI), the truth discovery divided attention-ResNet(TDA-ResNet)model are proposed based on improved residual network(ResNet)with truth discovery. Firstly, the ResNet network model architecture is optimized by the divided attention mechanism to extract local and global features of glioma images to predict the IDH1 gene status of glioma; meanwhile, the truth discovery algorithm is incorporated into the model to calibrate the uncertainty of the depth feature vectors as the prediction results, so as to improve the prediction accuracy of the model. The experimental data were collected from the MR images of some glioma patients in the Affiliated Hospital of Xuzhou Medical University and the public dataset of the cancer imaging archive(TCIA). The experimental accuracies of the TDA-ResNet model in the MR image dataset of glioma in the Affiliated Hospital of Xuzhou Medical University and the TCIA dataset were 95.73% and 94.3%. The experimental results show that the TDA-ResNet model can achieve non-invasive prediction and uncertainty calibration of IDH1 gene status of glioma, and its performance is better than the existing deep learning prediction model of IDH1 gene status, which is of great significance for the clinical diagnosis and treatment of glioma.

Key words: glioma, isocitrate dehydrogenase 1, deep learning, truth discovery, uncertainty calibration

中图分类号: 

  • TP391.41
[1] SLEDZINSKA P, BEBYN M, FURTAK J, et al. Current and promising treatment strategies in glioma[J]. Reviews in the Neurosciences, 2022, 34(5):483-516.
[2] LOUIS D N, PERRY A, WESSELING P, et al. The 2021 WHO classification of tumors of the central nervous system: a summary[J]. Neuro-Oncology, 2021, 23(8):1231-1251.
[3] XU Jie, XU Fangping, LIU Zhihua, et al. The correlation analysis of TERT promoter mutations with IDH1/2 mutations and 1p/19q detected in human gliomas[J]. Medicine, 2022, 101(29):e29668.
[4] OHBA S, KUWAHARA K, YAMADA S, et al. Correlation between IDH, ATRX, and TERT promoter mutations in glioma[J]. Brain Tumor Pathology, 2020, 37(2):33-40.
[5] LI Yiming, WEI Dong, LIU Xing, et al. Molecular subtyping of diffuse gliomas using magnetic resonance imaging: comparison and correlation between radiomics and deep learning[J]. European Radiology, 2022, 32(2):747-758.
[6] WENG G D, ERMI?瘙塁 E, MARAGKOU T, et al. Accurate prediction of isocitrate dehydrogenase-mutation status of gliomas using SLOW-editing magnetic resonance spectroscopic imaging at 7 T MR[J]. Neuro-Oncology Advances, 2023, 5(1):vdad001.
[7] XU Qian, XU Qianqian, SHI Nian, et al. A multitask classification framework based on vision transformer for predicting molecular expressions of glioma[J]. European Journal of Radiology, 2022, 157:110560
[8] MA Chunwei, HUANG Ziyun, XIAN Jiayi, et al. Improving uncertainty calibration of deep neural networks via truth discovery and geometric optimization[C] //Uncertainty in Artificial Intelligence. [S.l.] :PMLR, 2021:75-85.
[9] MURUGESAN B, LIU B Y, GALDRAN A, et al. Calibrating segmentation networks with margin-based label smoothing[J]. Medical Image Analysis, 2023, 87:102826.
[10] BUDDENKOTTE T, ESCUDERO SANCHEZ L, CRISPIN-ORTUZAR M, et al. Calibrating ensembles for scalable uncertainty quantification in deep learning-based medical image segmentation[J]. Computers in Biology and Medicine, 2023, 163:107096.
[11] CANT K, EVINS R. Improved calibration of building models using approximate Bayesian calibration and neural networks[J]. Journal of Building Performance Simulation, 2023, 16(3):291-307.
[12] SHAFIQ M, GU Z. Deep residual learning for image recognition: a survey[J]. Applied Sciences, 2022, 12(18):8972.
[13] XIE S, GIRSHICK R, DOLLÁR P, et al. Aggregated residual transformations for deep neural networks[C] //Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE, 2017:1492-1500.
[14] DING Hu, XU Jinhui. Learning the truth vector in high dimensions[J]. Journal of Computer and System Sciences, 2020, 109:78-94.
[15] KUMAR A, LIANG P, MA T. Verified uncertainty calibration[EB/OL].(2019-09-23)[2023-12-29]. http://arxiv.org/abs/1909.10155.
[16] GUO C, PLEISS G, SUN Y, et al. On calibration of modern neural networks[C] //Proceedings of the 34th International Conference on Machine Learning-Volume 70. Sydney: ACM, 2017:1321-1330.
[17] 曹建军,常宸,翁年凤,等. 基于神经网络编码的真值发现[J]. 计算机工程与科学,2021,43(9):1546-1557. CAO Jianjun, CHANG Chen, WENG Nianfeng, et al. Truth discovery based on neural network encoding[J]. Computer Engineering & Science, 2021, 43(9):1546-1557.
[18] XU Haitao, ZHANG Haiwang, LI Qianqian, et al. A data-semantic-conflict-based multi-truth discovery algorithm for a programming site[J]. Computers, Materials & Continua, 2021, 68(2):2681-2691.
[19] CHOI Y S, BAE S, CHANG J H, et al. Fully automated hybrid approach to predict the IDH mutation status of gliomas via deep learning and radiomics[J]. Neuro-Oncology, 2021, 23(2):304-313.
[20] TAHA B, LI T H, BOLEY D, et al. Detection of isocitrate dehydrogenase mutated glioblastomas through anomaly detection analytics[J]. Neurosurgery, 2021, 89(2):323-328.
[21] KAWAGUCHI R K, TAKAHASHI M, MIYAKE M, et al. Assessing versatile machine learning models for glioma radiogenomic studies across hospitals[J]. Cancers, 2021, 13(14):3611.
[1] 陈玉明,郑光宇,焦娜. 基于粒神经网络的多标签学习[J]. 《山东大学学报(理学版)》, 2024, 59(5): 1-11.
[2] 李程,车文刚,高盛祥. 一种用于航拍图像的目标检测算法[J]. 《山东大学学报(理学版)》, 2023, 58(9): 59-70.
[3] 徐华畅,许倩,赵钰琳,梁峰宁,徐凯,朱红. 基于改进EfficientNetV2的脑胶质瘤IDH1突变状态预测方法[J]. 《山东大学学报(理学版)》, 2023, 58(7): 60-66.
[4] 仲诚诚,周恒,张梓童,张春雷. LAC-UNet: 基于胶囊表达局部-整体特征关系的语义分割模型[J]. 《山东大学学报(理学版)》, 2023, 58(11): 116-126.
[5] 徐菲菲,许赟杰. 基于Arc-LSTM的人职匹配研究[J]. 《山东大学学报(理学版)》, 2021, 56(1): 83-90.
[6] 郝长盈,兰艳艳,张海楠,郭嘉丰,徐君,庞亮,程学旗. 基于拓展关键词信息的对话生成模型[J]. 《山东大学学报(理学版)》, 2019, 54(7): 68-76.
[7] 刘飚,路哲,黄雨薇,焦萌,李泉其,薛瑞. 神经网络结构在功耗分析中的性能对比[J]. 《山东大学学报(理学版)》, 2019, 54(1): 60-66.
[8] 庞博,刘远超. 融合pointwise及深度学习方法的篇章排序[J]. 山东大学学报（理学版）, 2018, 53(3): 30-35.
[9] 刘明明,张敏情,刘佳,高培贤. 一种基于浅层卷积神经网络的隐写分析方法[J]. 山东大学学报（理学版）, 2018, 53(3): 63-70.
[10] 刘铭, 昝红英, 原慧斌. 基于SVM与RNN的文本情感关键句判定与抽取[J]. 山东大学学报（理学版）, 2014, 49(11): 68-73.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2018 《山东大学学报(理学版)》编辑部　
地址: 山东省济南市山大南路27号山东大学中心校区（250100）　电话: +86-0531-88366917　E-mail: xblxb@sdu.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发