Brain age prediction using deep learning uncovers associated sequence variants.
B A JonssonGyda BjornsdottirThorgeir E ThorgeirssonLotta M EllingsenG Bragi WaltersDaníel F GuðbjartssonHreinn StefánssonKári StefánssonMagnus O UlfarssonPublished in: Nature communications (2019)
Machine learning algorithms can be trained to estimate age from brain structural MRI. The difference between an individual's predicted and chronological age, predicted age difference (PAD), is a phenotype of relevance to aging and brain disease. Here, we present a new deep learning approach to predict brain age from a T1-weighted MRI. The method was trained on a dataset of healthy Icelanders and tested on two datasets, IXI and UK Biobank, utilizing transfer learning to improve accuracy on new sites. A genome-wide association study (GWAS) of PAD in the UK Biobank data (discovery set: [Formula: see text], replication set: [Formula: see text]) yielded two sequence variants, rs1452628-T ([Formula: see text], [Formula: see text]) and rs2435204-G ([Formula: see text], [Formula: see text]). The former is near KCNK2 and correlates with reduced sulcal width, whereas the latter correlates with reduced white matter surface area and tags a well-known inversion at 17q21.31 (H2).
Keyphrases
- white matter
- deep learning
- machine learning
- smoking cessation
- human milk
- resting state
- contrast enhanced
- genome wide association study
- multiple sclerosis
- magnetic resonance imaging
- cerebral ischemia
- computed tomography
- convolutional neural network
- gene expression
- high throughput
- preterm infants
- body composition
- genome wide
- data analysis
- network analysis