微信服务号

微信订阅号

2025年6月1日 15:55 星期日
首页 > 过刊浏览>2024年第35卷第4期 >1682-1702. DOI:10.13328/j.cnki.jos.007012
PDF HTML阅读 XML下载 导出引用 引用提醒
神经网络结构搜索在脑数据分析领域的研究进展
作者:
通讯作者:

邬霞,E-mail:wuxia@bnu.edu.cn

基金项目:

国家自然科学基金(62206024, 62236001); 北京市自然科学基金(4212037)


Survey on Neural Architecture Search for Brain Data Analysis
Author:
  • 摘要
    • | |
  • 访问统计
    • |
  • 参考文献 [64]
    • |
  • 相似文献 [20]
    • | | |
  • 文章评论
    摘要:

    神经网络结构搜索(neural architecture search, NAS)是自动化机器学习的重要组成部分, 已被广泛应用于多个领域, 包括计算机视觉、语音识别等, 能够针对特定数据、场景、任务寻找最优的深层神经网络结构. 将NAS引入至脑数据分析领域, 能够在图像分割、特征提取、辅助诊断等多个应用领域大幅度提升性能, 展现低能耗自动化机器学习的优势. 基于NAS进行脑数据分析是当前的研究热点之一, 同时也具有一定挑战. 目前, 在此领域,国内外可供参考的综述性文献较少. 对近年来国内外相关文献进行了细致地调研分析, 从算法模型、研究任务、实验数据等不同方面对NAS在脑数据分析领域的研究现状进行了综述. 同时, 也对能够支撑NAS训练的脑数据集进行了系统性总结, 并对NAS在脑数据分析中存在的挑战和未来的研究方向进行了分析和展望.

    Abstract:

    Neural architecture search (NAS) is an important part of automated machine learning, which has been widely used in multiple fields, including computer vision, speech recognition, etc. NAS can search the optimal deep neural network structures for specific data, scenarios, and tasks. In recent years, NAS has been increasingly applied to brain data analysis, significantly improving the performance in multiple application fields, such as brain image segment, feature extraction, brain disease auxiliary diagnosis, etc. Such researches have demonstrated the advantages of low-energy automated machine learning in the field of brain data analysis. NAS-based brain data analysis is one of the current research hotspots, and it still has certain challenges. At present, there are few review literatures available for reference in this field worldwide. This study conducts a detailed survey and analysis of relevant literature from different perspectives, including search frameworks, search space, search strategies, research tasks, and experimental data. At the same time, a systematic summary of brain data sets is also provided that can be used for NAS training. In addition, challenges and future research directions of NAS are prospected in brain data analysis.

    参考文献
    [1] Tian JX, Liu GC, Gu SS, et al. Deep learning in medical image analysis and its challenges. Acta Automatica Sinica, 2018, 44(3):401-424 (in Chinese with English abstract)
    [2] Avberšek LK, Repovš G. Deep learning in neuroimaging data analysis:Applications, challenges, and solutions. Frontiers in Neuroimaging, 2022, 1:981642.
    [3] Li C, Zhang ZZ, Song RC, et al. EEG-based emotion recognition via neural architecture search. IEEE Trans. on Affective Computing, 2023:14:957-968.
    [4] Li Q, Dong QL, Ge FF, et al. Simultaneous spatial-temporal decomposition for connectome-scale brain networks by deep sparse recurrent auto-encoder. Brain Imaging and Behavior, 2021, 15:2646-2660.
    [5] Guo DZ, Jin DK, Zhu ZT, et al. Organ at risk segmentation for head and neck cancer using stratified learning and neural architecture search. In:Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR). 2020. 4222-4231.
    [6] Huang QY, Yang D, Xian YK, et al. Enhanced mri reconstruction network using neural architecture search. In:Proc. of the Int'l Workshop on Machine Learning in Medical Imaging-Machine Learning in Medical Imaging (MICCAI-MLMI). 2020. 634-643.
    [7] Elsken T, Metzen JH, Hutter F. Neural architecture search :A survey. Journal of Machine Learning Research, 2019, 20:1-21.
    [8] Li HY, Wang NN, Zhu MR, et al. Recent advances in neural architecture search:A survey. Ruan Jian Xue Bao/Journal of Software, 2021, 33(1):129-149 (in Chinese with English abstract). https://www.jos.org.cn/1000-9825/6306.htm[doi:10.13328/j.cnki.jos. 006306]
    [9] Liu SF, Ge FF, Zhao L, et al. NAS-optimized topology-preserving transfer learning for differentiating cortical folding patterns. Medical Image Analysis, 2022, 77:102316.
    [10] Li Q, Wu X, Liu TM. Differentiable neural architecture search for optimal spatial/temporal brain function network decomposition. Medical Image Analysis, 2021, 69:101974.
    [11] Ellermann J, Garwood M, Hendrich K, et al. Functional imaging of the brain by nuclear magnetic resonance. In:Proc. of the NMR in Physiology and Biomedicine. 1994. 137-150.
    [12] Schcter DL. EEG theta waves and psychological phenomena:a review and analysis. Biological Psychology, 1977, 5(1):47-82.
    [13] Madsen PL, Secher NH. Near-infrared oximetry of the brain. Progress in Neurobiology, 1999, 58(6):541-560.
    [14] Brown R, Cheng YC, Haacke EM, et al. Magnetic Resonance Imaging:Physical Principles and Sequence Design. Wiley, 1999.
    [15] Le Bihan D. Looking into the functional architecture of the brain with diffusion MRI. Nature Reviews Neuroscience, 2006, 1290:1-24.
    [16] Heinz ER, DuBois P, Osborne D, et al. Dynamic computed tomography study of the brain. Journal of Computer Assisted Tomography, 1979, 3(5):641-649.
    [17] Huang H, Hu XT, Zhao Y, et al. Modeling task fmri data via deep convolutional autoencoder. IEEE Trans. on Medical Imaging, 2018, 37(7):1551-1561.
    [18] Zhang W, Zhao SJ, Hu XT, et al. Hierarchical organization of functional brain networks revealed by hybrid spatiotemporal deep learning. Brain Connectivity, 2020, 10(2):72-82.
    [19] Li C, Zhang ZZ, Song RC, et al. EEG-based emotion recognition via transformer neural architecture search. IEEE Trans. on Industrial Informatics, 2023, 19(4):6016-6025.
    [20] Nguyen KP, Fatt CC, Treacher A, et al. Predicting response to the antidepressant bupropion using pretreatment fmri. In:Proc. of the Int'l Workshop on Machine Learning in Medical Imaging-PRedictive Intelligence in Medicine (MICCAI-PRIME), Vol.11843. 2019. 53-62.
    [21] Dai HX, Li Q, Zhao L, et al. Graph representation neural architecture search for optimal spatial/temporal functional brain network decomposition. In:Proc. of the Int'l Workshop on Machine Learning in Medical Imaging-Machine Learning in Medical Imaging (MICCAI-MLMI). 2022. 279-287.
    [22] Li Q, Zhang W, Zhao L, et al. Evolutional neural architecture search for optimization of spatiotemporal brain network decomposition. IEEE Trans. on Bio-medical Engineering, 2022, 69(2):624-634.
    [23] Liu HX, Simonyan K, Yang YM. DARTS:Differentiable architecture search. In:Proc. of the Int'l Conf. on Learning Representations (ICLR). 2019. 1-12.
    [24] Xiao AQ, Shen BL, Shi XJ, et al. Intraoperative glioma grading using neural architecture search and multi-modal imaging. IEEE Trans. on Medical Imaging, 2022, 41(10):2570-2581.
    [25] Rapaport E, Shriki O, Puzis R. EEGNAS:Neural architecture search for electroencephalography data analysis and decoding. In:Proc. of the Int'l Joint Conf. on Artificial Intelligence-human Brain and Artificial Intelligence (IJCAI-HBAI). 2019. 3-20.
    [26] Dong HY, Chen D, Zhang L, et al. Subject sensitive eeg discrimination with fast reconstructable cnn driven by reinforcement learning:A case study of asd evaluation. Neurocomputing, 2021, 449:136-145.
    [27] Du YP, Liu J, Wang X, et al. SSVEP-based emotion recognition for iot via multiobjective neural architecture search. IEEE Internet of Things Journal, 2022, 9(21):21432-21443.
    [28] Duan YQ, Wang Z, Li Y, et al. Cross task neural architecture search for EEG signal classifications. Neurocomputing, 2022, 545:126260.
    [29] Kong GW, Li C, Peng H, et al. EEG-based sleep stage classification via neural architecture search. IEEE Trans. on Neural Systems and Rehabilitation Engineering, 2023, 31:1075-1085.
    [30] Wang H, Zhu XS, Chen PY, et al. A gradient-based automatic optimization cnn framework for EEG state recognition. Journal of Neural Engineering, 2022, 19(1):016009.
    [31] Zhang W, Zhao L, Li Q, et al. Identify hierarchical structures from task-based FMRI data via hybrid spatiotemporal neural architecture search net. In:Proc. of the Medical Image Computing and Computer Assisted Intervention (MICCAI). 2019. 745-753.
    [32] Qiang N, Dong QL, Zhang W, et al. Modeling task-based fmri data via deep belief network with neural architecture search. In:Proc. of the Computerized Medical Imaging and Graphics. 2020. 101747.
    [33] Ren YD, Xu SH, Tao ZY, et al. Hierarchical spatio-temporal modeling of naturalistic functional magnetic resonance imaging signals via two-stage deep belief network with neural architecture search. Frontiers in Neuroscience, 2021, 15:794955.
    [34] Duan FX, Cao CH, Gao XP. SA-NAS-BFNR:Spatiotemporal attention neural architecture search for task-based brain functional network representation. In:Proc. of the Int'l Conf. on Multimedia Retrieval (ICMR). 2022. 661-667.
    [35] Pang TJ, Zhao SJ, Han JW, et al. Gumbel-softmax based neural architecture search for hierarchical brain networks decomposition. Medical Image Analysis, 2022, 82:102570.
    [36] Dai HX, Ge FF, Li Q, et al. Optimize CNN model for fmri signal classification via adanet-based neural architecture search. In:Proc. of the Int'l Symp. on Biomedical Imaging (ISBI). 2020. 1399-1403.
    [37] Cortes C, Gonzalvo X, Kuznetsov V, et al. AdaNet:Adaptive structural learning of artificial neural networks. In:Proc. of the Int'l Conf. on Machine Learning (PMLR), Vol.70. 2017. 874-883.
    [38] Sandler M, Howard A, Zhu ML, et al. MobileNetV2:Inverted residuals and linear bottlenecks. In:Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR). 2018. 4510-4520.
    [39] Xu YH, Xie LX, Zhang XP, et al. PC-DARTS:Partial channel connections for memory-efficient architecture search. In:Proc. of the Int'l Conf. on Learning Representations (ICLR). 2020. 1-13.
    [40] Bullmore E, Sporns O. Complex brain networks:Graph theoretical analysis of structural and functional systems. Nature Reviews Neuroscience, 2009, 10(3):186-198.
    [41] Kriegeskorte N, Douglas PK. Cognitive computational neuroscience. Nature Neuroscience, 2018, 21:1148-1160.
    [42] Xu SH, Ren YD, Tao ZY, et al. Hierarchical individual naturalistic functional brain networks with group consistency uncovered by a two-stage nas-volumetric sparse dbn framework. eNeuro, 2022, 9(5):1-11.
    [43] Li Q, Zhang W, Lv JL, et al. Neural architecture search for optimization of spatial-temporal brain network decomposition. In:Proc. of the Medical Image Computing and Computer Assisted Intervention (MICCAI). 2020. 1-10.
    [44] Tao ZY, Ren YD, Zhang W, et al. Identifying hierarchical individual functional network under naturalistic paradigm via two-stage dbn with neural architecture search. In:Proc. of the Int'l Symp. on Image Computing and Digital Medicine (ISICDM). 2021. 130-134.
    [45] Hu XB, Shen RL, Luo DH, et al. AutoGAN-synthesizer:Neural architecture search for cross-modality MRI synthesis. In:Proc. of the Medical Image Computing and Computer Assisted Intervention (MICCAI). 2022. 397-409.
    [46] Chen HZ, Zhang ZJ, Jin MW, et al. Prediction of dmri signals with neural architecture search. Journal of Neuroscience Methods, 2022, 365:109389.
    [47] Yan JP, Chen S, Zhang YB, et al. Neural architecture search for compressed sensing magnetic resonance image reconstruction. Computerized Medical Imaging and Graphics, 2020, 85:101784.
    [48] Eslahi SV, Tao J, Ji J. ERNAS:An evolutionary neural architecture search for magnetic resonance image reconstructions. arXiv:2206.07280, 2023.
    [49] Kim SW, Kim I, Lim SB, et al. Scalable neural architecture search for 3D medical image segmentation. In:Proc. of the Medical Image Computing and Computer Assisted Intervention (MICCAI). 2019. 220-228.
    [50] Wong KCL, Moradi M. SegNAS3D:Network architecture search with derivative-free global optimization for 3D image segmentation. In:Proc. of the Medical Image Computing and Computer Assisted Intervention (MICCAI). 2019. 393-401.
    [51] Wang FF. Neural architecture search for gliomas saegmentation on multimodal magnetic resonance imaging. arXiv:2005.06338, 2020.
    [52] He YF, Yang D, Roth H, et al. DiNTS:Differentiable neural network topology search for 3D medical image segmentation. IEEE/CVF Computer Vision and Pattern Recognition (CVPR). 2021. 5841-5850.
    [53] Milesi A, Futrega M, Marcinkiewicz M, et al. Brain tumor segmentation using neural network topology search. In:Proc. of the Brainlesion:Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries. 2022. 366-376.
    [54] Peng C, Myronenko A, Hatamizadeh A, et al. HyperSegNAS:Bridging one-shot neural architecture search with 3D medical image segmentation using hypernet. In:Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR). 2022. 20709-20719.
    [55] Xiang TG, Zhang CY, Wang XY, et al. Towards bi-directional skip connections in encoder-decoder architectures and beyond. Medical Image Analysis, 2022, 78:102420.
    [56] Ye XH, Guo DZ, Ge J, et al. Comprehensive and clinically accurate head and neck cancer organs-at-risk delineation on a multi-institutional study. Nature Communications, Nature Publishing Group, 2022, 13(1):6137.
    [57] Bae W, Lee SH, Lee YH, et al. Resource optimized neural architecture search for 3D medical image segmentation. In:Proc. of the Medical Image Computing and Computer Assisted Intervention (MICCAI). 2019. 228-236.
    [58] Yu QH, Yang D, Roth H, et al. C2FNAS:Coarse-to-fine neural architecture search for 3D medical image segmentation. In:Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition (CVPR). 2020. 4125-4134.
    [59] Calisto MB, Lai-Yuen SK. EMONAS-Net:Efficient multiobjective neural architecture search using surrogate-assisted evolutionary algorithm for 3D medical image segmentation. Artificial Intelligence in Medicine, 2021, 119:102154.
    [60] Tchetchenian A, Zhu YM, Zhang F, et al. A comparison of manual and automated neural architecture search for white matter tract segmentation. Scientific Reports, 2023, 13(1):1617.
    [61] Chitnis S, Hosseini R, Xie PT. Brain tumor classification based on neural architecture search. Scientific Reports, 2022, 12(1):19206.
    附中文参考文献:
    [1] 田娟秀, 刘国才, 谷珊珊, 等. 医学图像分析深度学习方法研究与挑战. 自动化学报, 2018, 44(3):401-424.
    [8] 李航宇, 王楠楠, 朱明瑞, 等. 神经结构搜索的研究进展综述. 软件学报, 2021, 33(1):129-149. https://www.jos.org.cn/1000-9825/6306.htm[doi:10.13328/j.cnki.jos.006306]
    引证文献
    网友评论
    网友评论
    分享到微博
    发 布
引用本文

李晴,汪启昕,李子遇,祝志远,张诗皓,牟浩南,杨文婷,邬霞.神经网络结构搜索在脑数据分析领域的研究进展.软件学报,2024,35(4):1682-1702

复制
分享
文章指标
  • 点击次数:1248
  • 下载次数: 4080
  • HTML阅读次数: 2240
  • 引用次数: 0
历史
  • 收稿日期:2023-05-14
  • 最后修改日期:2023-07-07
  • 在线发布日期: 2023-09-11
  • 出版日期: 2024-04-06
文章二维码
友情链接:中国科学院软件研究所中国计算机学会中国科学院国家自然科学基金委员会CCF数字图书馆Int'l J of Softw Info计算机系统应用
您是第19987115位访问者
版权所有：中国科学院软件研究所 京ICP备05046678号-3
地址：北京市海淀区中关村南四街4号,邮政编码：100190
电话：010-62562563 传真：010-62562533 Email：jos@iscas.ac.cn
技术支持：北京勤云科技发展有限公司

京公网安备 11040202500063号