A practical Alzheimer's disease classifier via brain imaging-based deep learning on 85,721 samples

doi:10.1186/s40537-022-00650-y

PSYCH OpenIR

	A practical Alzheimer's disease classifier via brain imaging-based deep learning on 85,721 samples
	Lu, Bin 1,2; Li, Hui-Xian 1,2; Chang, Zhi-Kai 1,2; Li, Le 12; Chen, Ning-Xuan 1,2; Zhu, Zhi-Chen 1,2; Zhou, Hui-Xia 1,2,3; Li, Xue-Ying 1,4,13; Wang, Yu-Wei 1,2; Cui, Shi-Xian 1,4,13; Deng, Zhao-Yu 1,2; Fan, Zhen 5; Yang, Hong 6; Chen, Xiao 1,2; Thompson, Paul M.7; Castellanos, Francisco Xavier 8,9; Yan, Chao-Gan 1,2,10,11
通讯作者	Yan, Chao-Gan(yancg@psych.ac.cn)
摘要	Beyond detecting brain lesions or tumors, comparatively little success has been attained in identifying brain disorders such as Alzheimer's disease (AD), based on magnetic resonance imaging (MRI). Many machine learning algorithms to detect AD have been trained using limited training data, meaning they often generalize poorly when applied to scans from previously unseen scanners/populations. Therefore, we built a practical brain MRI-based AD diagnostic classifier using deep learning/transfer learning on a dataset of unprecedented size and diversity. A retrospective MRI dataset pooled from more than 217 sites/scanners constituted one of the largest brain MRI samples to date (85,721 scans from 50,876 participants) between January 2017 and August 2021. Next, a state-of-the-art deep convolutional neural network, Inception-ResNet-V2, was built as a sex classifier with high generalization capability. The sex classifier achieved 94.9% accuracy and served as a base model in transfer learning for the objective diagnosis of AD. After transfer learning, the model fine-tuned for AD classification achieved 90.9% accuracy in leave-sites-out cross-validation on the Alzheimer's Disease Neuroimaging Initiative (ADNI, 6,857 samples) dataset and 94.5%/93.6%/91.1% accuracy for direct tests on three unseen independent datasets (AIBL, 669 samples / MIRIAD, 644 samples / OASIS, 1,123 samples). When this AD classifier was tested on brain images from unseen mild cognitive impairment (MCI) patients, MCI patients who converted to AD were 3 times more likely to be predicted as AD than MCI patients who did not convert (65.2% vs. 20.6%). Predicted scores from the AD classifier showed significant correlations with illness severity. In sum, the proposed AD classifier offers a medical-grade marker that has potential to be integrated into AD diagnostic practice.
关键词	Alzheimer's disease Convolutional neural network Magnetic resonance brain imaging Sex differences Transfer learning
	2022-10-13
语种	英语
DOI	10.1186/s40537-022-00650-y
发表期刊	JOURNAL OF BIG DATA
卷号	9 期号:1 页码:22
收录类别	SCI
资助项目	Sci-Tech Innovation 2030 -Major Project of Brain Science and Brain-inspired Intelligence Technology[2021ZD0200600] ; National Key R&D Program of China[2017YFC1309902] ; National Natural Science Foundation of China[82122035] ; National Natural Science Foundation of China[81671774] ; National Natural Science Foundation of China[81630031] ; 13th Five-year Informatization Plan of Chinese Academy of Sciences[XXH13505] ; Key Research Program of the Chinese Academy of Sciences[ZDBS-SSW-JSC006] ; Beijing Nova Program of Science and Technology[Z191100001119104] ; Scientific Foundation of Institute of Psychology, Chinese Academy of Sciences[E2CX4425YZ]
出版者	SPRINGERNATURE
WOS关键词	MRI ; SEX ; AGE ; REVEALS ; NETWORK ; CANCER
WOS研究方向	Computer Science
WOS类目	Computer Science, Theory & Methods
WOS记录号	WOS:000867657800001
资助机构	Sci-Tech Innovation 2030 -Major Project of Brain Science and Brain-inspired Intelligence Technology ; National Key R&D Program of China ; National Natural Science Foundation of China ; 13th Five-year Informatization Plan of Chinese Academy of Sciences ; Key Research Program of the Chinese Academy of Sciences ; Beijing Nova Program of Science and Technology ; Scientific Foundation of Institute of Psychology, Chinese Academy of Sciences
引用统计
文献类型	期刊论文
条目标识符	https://ir.psych.ac.cn/handle/311026/43506
通讯作者	Yan, Chao-Gan
作者单位	1.Inst Psychol, CAS Key Lab Behav Sci, Beijing, Peoples R China 2.Univ Chinese Acad Sci, Dept Psychol, Beijing, Peoples R China 3.Inst Psychol, CAS Key Lab Mental Hlth, Beijing, Peoples R China 4.Sino Danish Ctr Educ & Res, Beijing, Peoples R China 5.Fudan Univ, Huashan Hosp, Dept Neurosurg, Shanghai, Peoples R China 6.Zhejiang Univ, Coll Med, Affiliated Hosp 1, Dept Radiol, Hangzhou, Zhejiang, Peoples R China 7.Univ Southern Calif, Keck Sch Med, Mark & Mary Stevens Inst Neuroimaging & Informat, Imaging Genet Ctr, Los Angeles, CA 90007 USA 8.NYU Grossman Sch Med, Dept Child & Adolescent Psychiat, New York, NY USA 9.Nathan S Kline Inst Psychiat Res, Orangeburg, NY USA 10.Chinese Acad Sci, Inst Psychol, Int Big Data Ctr Depress Res, Beijing, Peoples R China 11.Chinese Acad Sci, Magnet Resonance Imaging Res Ctr, Inst Psychol, Beijing, Peoples R China 12.Beijing Language & Culture Univ, Ctr Cognit Sci Language, Beijing, Peoples R China 13.Univ Chinese Acad Sci, Sino Danish Coll, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Lu, Bin,Li, Hui-Xian,Chang, Zhi-Kai,et al. A practical Alzheimer's disease classifier via brain imaging-based deep learning on 85,721 samples[J]. JOURNAL OF BIG DATA,2022,9(1):22.
APA	Lu, Bin.,Li, Hui-Xian.,Chang, Zhi-Kai.,Li, Le.,Chen, Ning-Xuan.,...&Yan, Chao-Gan.(2022).A practical Alzheimer's disease classifier via brain imaging-based deep learning on 85,721 samples.JOURNAL OF BIG DATA,9(1),22.
MLA	Lu, Bin,et al."A practical Alzheimer's disease classifier via brain imaging-based deep learning on 85,721 samples".JOURNAL OF BIG DATA 9.1(2022):22.