第4期 孔伶旭，等：迁移学习特征提取的s-MRI早期轻度认知障碍分类 ·671· 该结果说明，ImageNet数据库中预训练的迁移网 [12]RUSSAKOVSKY O,DENG J,SU H,et al.ImageNet 络从ROI提取特征的分类准确率要高于从脑区网 large scale visual recognition challenge[J].International 络的功能性连接提取特征的方法，并且这种方法 Journal of Computer Vision,2015,115(3):211-252. 还可减少特征提取的时间和分类网络运行时间。 [13]BACKMAN L,JONES S,BERGER A K,et al.Multiple 参考文献： cognitive deficits during the transition to Alzheimer's dis- ease[J].Journal of internal medicine,2004,256(3): [1]BURNS A.ILIFFE S.Alzheimer's disease[J].British med- 195-204 ical journal,.2009,338(7692):467-471 [14]WEE C Y,YAP P T,DENNY K,et al.Resting-state [2]Alzheimer's Association.2018 Alzheimer's disease facts multi-spectrum functional connectivity networks for iden- and figures[J].Alzheimer's dementia,2018,14(3): tification of MCI patients[J].PLoS One,2012,7(5) 367-429. e37828. [3]DAVATZIKOS C.Baseline and longitudinal patterns of [15]GEHRING J,MIAO Y J,METZE F,et al.Extracting brain atrophy in MCI patients and their use in prediction of deep bottleneck features using stacked auto- short-term conversion to Alzheimer's disease:results from encoders[C]//Proceedings of 2013 IEEE International ADNI[J].Alzheimer's dementia,2009,5(4S):P21-P22. Conference on Acoustics,Speech and Signal Processing [4]GRUNDMAN M,PETERSEN R C,FERRIS S H.Mild Vancouver,BC,Canada,2013:3377-3381 cognitive impairment can be distinguished from Alzheimer [16]KRIZHEVSKY A,SUTSKEVER I,HINTON G E.Im- disease and normal aging for clinical trials[J].Archives of ageNet classification with deep convolutional neural net- neurology,2004,61(1):59-66. works[J].Communications of the ACM,2017,60(6): [5]BISCHKOPF J.BUSSE A.ANGERMEYER M C.Mild 84-90. cognitive impairment-a review of prevalence,incidence [17]HON M.KHAN N M.Towards Alzheimer's disease clas- and outcome according to current approaches[J].Acta psy- sification through transfer learning[C]//Proceedings of chiatrica scandinavica.2002.106(6):403-414. 2017 IEEE International Conference on Bioinformatics [6]SHARMA N,SINGH A N.Exploring biomarkers for and Biomedicine.Kansas City,MO,USA,2017: Alzheimer's disease[J].Journal of clinical and diagnostic 1166-1169. research,2016,10(7):KE01-KE06. [18]JOHNSON K A,FOX N C,SPERLING R A,et al.Brain [7]MCKAHNN G.Clinical diagnosis of Alzheimer's disease: Imaging in Alzheimer Disease[J].Cold spring harbor per- report of the NINCDS-ADRDA Work Group*under the spectives in medicine,2012,2(4):a006213. auspices of department of health and human services task [19]CUINGNET R.GERARDIN E.TESSIERAS J.et al force on Alzheimer's disease[J].Neurology,1984,34(7): Automatic classification of patients with Alzheimer's dis- 939. ease from structural MRI:a comparison of ten methods [8]SCHROETER ML.STEIN T.MASLOWSKI N.et al. using the ADNI database[J].Neuroimage,2011,56(2): Neural correlates of Alzheimer's disease and mild cognit- 766-781」 ive impairment:a systematic and quantitative meta-analys- [20]BARKHOF F,HALLER S,ROMBOUTS S A R B.Rest- is involving 1351 patients[J].Neurolmage,2009,47(4): ing-state functional MR imaging:a new window to the 1196-1206 brain[J].Radiology,2014,272(1):29-49. [9]HALLER S,BARTSCH A J.Pitfalls in fMRI[J].European [21]MACHULDA MM,WARD H A,BOROWSKI B,et al. radiology,.2009,1911)2689-2706. Comparison of memory fMRI response among Normal [10]BRIER M R,THOMAS J B,SNYDER A Z,et al.Loss of MCI,and Alzheimer's patients[J].Neurology,2003, intranetwork and internetwork resting state functional 61(4:500-506 connections with Alzheimer's disease progression[J]. [22]JU Ronghui,HU Chenhui,ZHOU Pan,et al.Early dia- Journal of neuroscience,2012,32(26):8890-8899 gnosis of Alzheimer's disease based on resting-state brain [11]STANLEY ML.MOUSSA MN.PAOLINI B M.et al. networks and deep learning[J].IEEE/ACM Transactions Defining nodes in complex brain networks[J].Frontiers in on Computational Biology and Bioinformatics,2019, computational neuroscience,2013,7:169 16(1244-257.该结果说明，ImageNet 数据库中预训练的迁移网 络从 ROI 提取特征的分类准确率要高于从脑区网 络的功能性连接提取特征的方法，并且这种方法 还可减少特征提取的时间和分类网络运行时间。 参考文献： BURNS A, ILIFFE S. Alzheimer’s disease[J]. British med￾ical journal, 2009, 338(7692): 467–471. [1] Alzheimer’s Association. 2018 Alzheimer’s disease facts and figures[J]. Alzheimer’s & dementia, 2018, 14(3): 367–429. [2] DAVATZIKOS C. Baseline and longitudinal patterns of brain atrophy in MCI patients and their use in prediction of short-term conversion to Alzheimer’s disease: results from ADNI[J]. Alzheimer’s & dementia, 2009, 5(4S): P21–P22. [3] GRUNDMAN M, PETERSEN R C, FERRIS S H. Mild cognitive impairment can be distinguished from Alzheimer disease and normal aging for clinical trials[J]. Archives of neurology, 2004, 61(1): 59–66. [4] BISCHKOPF J, BUSSE A, ANGERMEYER M C. Mild cognitive impairment-a review of prevalence, incidence and outcome according to current approaches[J]. Acta psy￾chiatrica scandinavica, 2002, 106(6): 403–414. [5] SHARMA N, SINGH A N. Exploring biomarkers for Alzheimer's disease[J]. Journal of clinical and diagnostic research, 2016, 10(7): KE01–KE06. [6] MCKAHNN G. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA Work Group* under the auspices of department of health and human services task force on Alzheimer's disease[J]. Neurology, 1984, 34(7): 939. [7] SCHROETER M L, STEIN T, MASLOWSKI N, et al. Neural correlates of Alzheimer’s disease and mild cognit￾ive impairment: a systematic and quantitative meta-analys￾is involving 1351 patients[J]. NeuroImage, 2009, 47(4): 1196–1206. [8] HALLER S, BARTSCH A J. Pitfalls in fMRI[J]. European radiology, 2009, 19(11): 2689–2706. [9] BRIER M R, THOMAS J B, SNYDER A Z, et al. Loss of intranetwork and internetwork resting state functional connections with Alzheimer’s disease progression[J]. Journal of neuroscience, 2012, 32(26): 8890–8899. [10] STANLEY M L, MOUSSA M N, PAOLINI B M, et al. Defining nodes in complex brain networks[J]. Frontiers in computational neuroscience, 2013, 7: 169. [11] RUSSAKOVSKY O, DENG J, SU H, et al. ImageNet large scale visual recognition challenge[J]. International Journal of Computer Vision, 2015, 115(3): 211–252. [12] BÄCKMAN L, JONES S, BERGER A K, et al. Multiple cognitive deficits during the transition to Alzheimer's dis￾ease[J]. Journal of internal medicine, 2004, 256(3): 195–204. [13] WEE C Y, YAP P T, DENNY K, et al. Resting-state multi-spectrum functional connectivity networks for iden￾tification of MCI patients[J]. PLoS One, 2012, 7(5): e37828. [14] GEHRING J, MIAO Y J, METZE F, et al. Extracting deep bottleneck features using stacked auto￾encoders[C]//Proceedings of 2013 IEEE International Conference on Acoustics, Speech and Signal Processing. Vancouver, BC, Canada, 2013: 3377−3381. [15] KRIZHEVSKY A, SUTSKEVER I, HINTON G E. Im￾ageNet classification with deep convolutional neural net￾works[J]. Communications of the ACM, 2017, 60(6): 84–90. [16] HON M, KHAN N M. Towards Alzheimer's disease clas￾sification through transfer learning[C]//Proceedings of 2017 IEEE International Conference on Bioinformatics and Biomedicine. Kansas City, MO, USA, 2017: 1166−1169. [17] JOHNSON K A, FOX N C, SPERLING R A, et al. Brain Imaging in Alzheimer Disease[J]. Cold spring harbor per￾spectives in medicine, 2012, 2(4): a006213. [18] CUINGNET R, GERARDIN E, TESSIERAS J, et al. Automatic classification of patients with Alzheimer’s dis￾ease from structural MRI: a comparison of ten methods using the ADNI database[J]. Neuroimage, 2011, 56(2): 766–781. [19] BARKHOF F, HALLER S, ROMBOUTS S A R B. Rest￾ing-state functional MR imaging: a new window to the brain[J]. Radiology, 2014, 272(1): 29–49. [20] MACHULDA M M, WARD H A, BOROWSKI B, et al. Comparison of memory fMRI response among Normal, MCI, and Alzheimer's patients[J]. Neurology, 2003, 61(4): 500–506. [21] JU Ronghui, HU Chenhui, ZHOU Pan, et al. Early dia￾gnosis of Alzheimer’s disease based on resting-state brain networks and deep learning[J]. IEEE/ACM Transactions on Computational Biology and Bioinformatics, 2019, 16(1): 244–257. [22] 第 4 期 孔伶旭，等：迁移学习特征提取的 rs-fMRI 早期轻度认知障碍分类 ·671·