Deep transfer learning of structural magnetic resonance imaging fused with blood parameters improves brain age prediction
Open Access
- 16 December 2021
- journal article
- research article
- Published by Wiley in Human Brain Mapping
- Vol. 43 (5) , 1640-1656
Abstract
Machine learning has been applied to neuroimaging data for estimating brain age and capturing early cognitive impairment in neurodegenerative diseases. Blood parameters like neurofilament light chain are associated with aging. In order to improve brain age predictive accuracy, we constructed a model based on both brain structural magnetic resonance imaging (sMRI) and blood parameters. Healthy subjects (n = 93; 37 males; aged 50–85 years) were recruited. A deep learning network was firstly pretrained on a large set of MRI scans (n = 1,481; 659 males; aged 50–85 years) downloaded from multiple open‐source datasets, to provide weights on our recruited dataset. Evaluating the network on the recruited dataset resulted in mean absolute error (MAE) of 4.91 years and a high correlation (r = .67, p <.001) against chronological age. The sMRI data were then combined with five blood biochemical indicators including GLU, TG, TC, ApoA1 and ApoB, and 9 dementia‐associated biomarkers including ApoE genotype, HCY, NFL, TREM2, Aβ40, Aβ42, T‐tau, TIMP1, and VLDLR to construct a bilinear fusion model, which achieved a more accurate prediction of brain age (MAE, 3.96 years; r = .76, p <.001). Notably, the fusion model achieved better improvement in the group of older subjects (70–85 years). Extracted attention maps of the network showed that amygdala, pallidum, and olfactory were effective for age estimation. Mediation analysis further showed that brain structural features and blood parameters provided independent and significant impact. The constructed age prediction model may have promising potential in evaluation of brain health based on MRI and blood parameters.Keywords
Funding Information
- National Natural Science Foundation of China (21877081)
- Science, Technology and Innovation Commission of Shenzhen Municipality (JCYJ20200109110001818)
- Action Medical Research
- Medical Research Council
- Engineering and Physical Sciences Research Council
- H2020 European Research Council
- National Institute of Biomedical Imaging and Bioengineering
- National Institute on Aging
- National Institutes of Health (R01 EB009352, UL1 TR000448, R01 AG043434, P01 AG003991, P01 AG026276, P30 NS09857781, P50 AG00561, U01 AG024904)
- Alzheimer's Disease Neuroimaging Initiative
This publication has 55 references indexed in Scilit:
- Deep biomarkers of human aging: Application of deep neural networks to biomarker developmentAging, 2016
- Opposing effects ofApoe/Apoa1double deletion on amyloid-β pathology and cognitive performance in APP miceBrain, 2015
- APOE-by-sex interactions on brain structure and metabolism in healthy elderly controlsOncotarget, 2015
- Brain structure and function as mediators of the effects of amyloid on memoryNeurology, 2015
- Cortical and Subcortical Atrophy in Alzheimer DiseaseAlzheimer Disease & Associated Disorders, 2014
- Clinical and Biomarker Changes in Dominantly Inherited Alzheimer's DiseaseNew England Journal of Medicine, 2012
- Plasma Amyloid-β as a Predictor of Dementia and Cognitive DeclineArchives of Neurology, 2012
- A general approach to causal mediation analysis.Psychological Methods, 2010
- On the Adaptive Control of the False Discovery Rate in Multiple Testing With Independent StatisticsJournal of Educational and Behavioral Statistics, 2000
- Cholesterol metabolism and agingThe American Journal of Medicine, 1987