TY - JOUR
T1 - On the design of convolutional neural networks for automatic detection of Alzheimer's disease
AU - Liu, Sheng
AU - Yadav, Chhavi
AU - Fernandez-Granda, Carlos
AU - Razavian, Narges
N1 - Funding Information:
The authors would like to thank Henry Rusinek and Arjun Masurkar for their useful comments on earlier versions of this manuscript. Authors also acknowledge Leon Lowenstein Foundation for funding support, and the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (National Institutes of Health Grant U01 AG024904) and DOD ADNI (Department of Defense award number W81XWH-12-2-0012), as well as National Institute on Aging and the National Institute of Biomedical Imaging and Bioengineering for data collection and sharing.
Publisher Copyright:
© 2020 S. Liu, C. Yadav, C. Fernandez-Granda & N. Razavian.
PY - 2019
Y1 - 2019
N2 - Early detection is a crucial goal in the study of Alzheimer's Disease (AD). In this work, we describe several techniques to boost the performance of 3D convolutional neural networks trained to detect AD using structural brain MRI scans. Specifically, we provide evidence that (1) instance normalization outperforms batch normalization, (2) early spatial downsampling negatively affects performance, (3) widening the model brings consistent gains while increasing the depth does not, and (4) incorporating age information yields moderate improvement. Together, these insights yield an increment of approximately 14% in test accuracy over existing models when distinguishing between patients with AD, mild cognitive impairment, and controls in the ADNI dataset. Similar performance is achieved on an independent dataset.
AB - Early detection is a crucial goal in the study of Alzheimer's Disease (AD). In this work, we describe several techniques to boost the performance of 3D convolutional neural networks trained to detect AD using structural brain MRI scans. Specifically, we provide evidence that (1) instance normalization outperforms batch normalization, (2) early spatial downsampling negatively affects performance, (3) widening the model brings consistent gains while increasing the depth does not, and (4) incorporating age information yields moderate improvement. Together, these insights yield an increment of approximately 14% in test accuracy over existing models when distinguishing between patients with AD, mild cognitive impairment, and controls in the ADNI dataset. Similar performance is achieved on an independent dataset.
UR - http://www.scopus.com/inward/record.url?scp=85161614893&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85161614893&partnerID=8YFLogxK
M3 - Conference article
AN - SCOPUS:85161614893
SN - 2640-3498
VL - 116
SP - 184
EP - 201
JO - Proceedings of Machine Learning Research
JF - Proceedings of Machine Learning Research
T2 - Machine Learning for Health Workshop, ML4H 2019
Y2 - 13 December 2019
ER -