Accurate brain age prediction with lightweight deep neural networks

Abstract

Deep learning has huge potential for accurate disease prediction with neuroimaging data, but the prediction performance is often limited by training-dataset size and computing memory requirements. To address this, we propose a deep convolutional neural network model, Simple Fully Convolutional Network (SFCN), for accurate prediction of brain age using T1-weighted structural MRI data. Compared with other popular deep network architectures, SFCN has fewer parameters, so is more compatible with small dataset size and 3D volume data. The network architecture was combined with several techniques for boosting performance, including data augmentation, pre-training, model regularization, model ensemble and prediction bias correction. We compared our overall SFCN approach with several widely-used machine learning models. It achieved state-of-the-art performance in UK Biobank data (N = 14,503), with mean absolute error (MAE) = 2.14y in brain age prediction and 99.5% in sex classification. SFCN also won (both parts of) the 2019 Predictive Analysis Challenge for brain age prediction, involving 79 competing teams (N = 2,638, MAE = 2.90y). We describe here the details of our approach, and its optimisation and validation. Our approach can easily be generalised to other tasks using different image modalities, and is released on GitHub. Highlights A lightweight deep learning model, Simple Fully Convolutional Network (SFCN), is presented, achieving state-of-the-art brain age prediction and sex classification performance in UK Biobank MRI brain imaging data. Even with limited number of training subjects (e.g., 50), SFCN performs better than widely-used regression models. A semi-multimodal ensemble strategy is proposed and achieved first place in the PAC 2019 brain age prediction challenge. Linear regression can remove brain age prediction bias (even on unlabelled data) while maintaining state-of-the-art performance.

References 54

1. 1.
   Image processing and Quality Control for the first 10,000 brain imaging datasets from UK Biobank · 2018

   No notice
2. 2.

   Arslan S, Ktena SI, Glocker B, Rueckert D. Graph saliency maps through spectral convolutional networks: application to sex classification with brain connectivity. Graphs in Biomedical Image Analysis and Integrating Medical Imaging and Non-Imaging Modalities. 2018;3-13.

   Not in our library

3. 3.

   Ashburner J. A fast diffeomorphic image registration algorithm. Neuroimage. 2007;38:95-113.

   Not in our library10.1016/j.neuroimage.2007.07.007

4. 4.

   Bashyam VM, Erus G, Doshi J, Habes M, Nasralah I, Truelove-Hill M, Srinivasan D, Mamourian L, Pomponio R, Fan Y, Launer LJ, et al. MRI signatures of brain age and disease over the lifespan based on a deep brain network and 14 468 individuals worldwide. Brain. 2020.

   Not in our library10.1093/brain/awaa160

5. 5.

   Baumgartner CF, Koch LM, Tezcan KC, Ang JX. Visual feature attribution using wasserstein GANs. 2018;8309-8319.

   Not in our library

6. 6.

   Brown TT, Kuperman JM, Chung Y, Erhart M, McCabe C, Hagler DJ, Venkatraman VK, Akshoomoff N, Amaral DG, Bloss CS, Casey BJ, et al. Neuroanatomical assessment of biological maturity. Curr Biol. 2012;22:1693-1698.

   Not in our library10.1016/j.cub.2012.07.002

7. 7.

   Brügger R, Baumgartner CF, Konukoglu E. A Partially Reversible U-Net for Memory-Efficient Volumetric Image Segmentation. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 2019;429-437.

   Not in our library

8. 8.

   Chekroud AM, Ward EJ, Rosenberg MD, Holmes AJ. Patterns in the human brain mosaic discriminate males from females. Proc Natl Acad Sci U S A. 2016.

   Not in our library10.1073/pnas.1523888113

9. 9.

   Cole JH, Franke K. Predicting age using neuroimaging: innovative brain ageing biomarkers. Trends Neurosci. 2017;40:681-690.

   Not in our library10.1016/j.tins.2017.10.001

10. 10.

    Cole JH, Poudel RPK, Tsagkrasoulis D, Caan MWA, Steves C, Spector TD, Montana G. NeuroImage Predicting brain age with deep learning from raw imaging data results in a reliable and heritable biomarker. Neuroimage. 2017;163:115-124.

    Not in our library10.1016/j.neuroimage.2017.07.059

11. 11.
    Brain age predicts mortality · 2017

    No notice
12. 12.

    Davatzikos C, Xu F, An Y, Fan Y, Resnick SM. Longitudinal progression of Alzheimer’s-like patterns of atrophy in normal older adults: the SPARE-AD index. Brain. 2009;132:2026-2035.

    Not in our library10.1093/brain/awp091

13. 13.

    De Fauw J, Ledsam JR, Romera-Paredes B, Nikolov S, Tomasev N, Blackwell S, Askham H, Glorot X, O’Donoghue B, Visentin D, van den Driessche G, et al. Clinically applicable deep learning for diagnosis and referral in retinal disease. Nat Med. 2018;24:1342-1350.

    Not in our library10.1038/s41591-018-0107-6

14. 14.

    Feng X, Lipton ZC, Yang J, Small SA, Provenzano FA. Estimating brain age based on a healthy population with deep learning and structural MRI arXiv Prepr. arXiv1907.00943. 2019.

    Not in our library10.1016/j.neurobiolaging.2020.02.009

15. 15.
    Ten Years of BrainAGE as a Neuroimaging Biomarker of Brain Aging: What Insights Have We Gained? · 2019

    No notice
16. 16.

    Franke K, Ristow M, Gaser C. Gender-specific impact of personal health parameters on individual brain aging in cognitively unimpaired elderly subjects. Front Aging Neurosci. 2014;6:94.

    Not in our library10.3389/fnagi.2014.00094

17. 17.

    Franke K, Ziegler G, Klöppel S, Gaser C. Estimating the age of healthy subjects from T1-weighted MRI scans using kernel methods: Exploring the influence of various parameters. Neuroimage. 2010;50:883-892.

    Not in our library10.1016/j.neuroimage.2010.01.005

18. 18.

    Friedman J, Hastie T, Tibshirani R. Regularization paths for generalized linear models via coordinate descent. J Stat Softw. 2010;33:1-22.

    Not in our library10.18637/jss.v033.i01

19. 19.

    Giudice M, Del Lippa RA, Puts DA, Bailey DH, Bailey JM, Schmitt DP. Joel et al.’s method systematically fails to detect large, consistent sex differences. Proc Natl Acad Sci U S A. 2016.

    Not in our library10.1073/pnas.1525534113

20. 20.

    Habes M, Erus G, Toledo JB, Zhang T, Bryan N, Launer LJ, Rosseel Y, Janowitz D, Doshi J, Van der Auwera S, von Sarnowski B, et al. White matter hyperintensities and imaging patterns of brain ageing in the general population. Brain. 2016;139:1164-1179.

    Not in our library10.1093/brain/aww008

21. 21.

    He K, Zhang X, Ren S, Sun J. Deep Residual Learning for Image Recognition. 2016;770-778.

    Not in our library

22. 22.

    He T, Kong R, Holmes AJ, Nguyen M, Sabuncu MR, Eickhoff SB, Bzdok D, Feng J, Yeo BTT. Deep neural networks and kernel regression achieve comparable accuracies for functional connectivity prediction of behavior and demographics. Neuroimage. 2020;206:116276.

    Not in our library10.1016/_jneuroimage.2019.116276

23. 23.

    Ioffe S, Szegedy C. Batch normalization: accelerating deep network training by reducing internal covariate shift. 2015;448-456.

    Not in our library

24. 24.

    Joel D, Berman Z, Tavor I, Wexler N, Gaber O, Stein Y, Shefi N, Pool J, Urchs S, Margulies DS, Liem F, et al. Sex beyond the genitalia: The human brain mosaic. Proc Natl Acad Sci U S A. 2015;112:15468-15473.

    Not in our library10.1073/pnas.1509654112

25. 25.

    Joel D, Persico A, Hänggi J, Pool J, Berman Z. Do brains of females and males belong to two distinct populations? Proc Natl Acad Sci U S A. 2016.

    Not in our library10.1073/pnas.1600792113

26. 26.

    Jollans L, Boyle R, Artiges E, Banaschewski T, Desrivières S, Grigis A, Mar-tinot JL, Paus T, Smolka MN, Walter H, Schumann G, et al. Quantifying performance of machine learning methods for neuroimaging data. Neuroimage. 2019;199:351-365.

    Not in our library10.1016/jneuroimage.2019.05.082

27. 27.
    Brain age prediction using deep learning uncovers associated sequence variants · 2019

    No notice
28. 28.

    Kamnitsas K, Ledig C, Newcombe VFJ, Simpson JP, Kane AD, Menon DK, Rueckert D, Glocker B. Efficient multi-scale 3D CNN with fully connected CRF for accurate brain lesion segmentation. Med Image Anal. 2017;36:61-78.

    Not in our library10.1016/jmedia.2016.10.004

29. 29.

    Kaufmann T, van der Meer D, Doan NT, Schwarz E, Lund MJ, Agartz I, Al-næs D, Barch DM, Baur-Streubel R, Bertolino A, Bettella F, et al. Common brain disorders are associated with heritable patterns of apparent aging of the brain. Nat Neurosci. 2019;22:1617-1623.

    Not in our library10.1038/s41593-019-0471-7

30. 30.

    Kawahara J, Brown CJ, Miller SP, Booth BG, Chau V, Grunau RE, Zwicker JG, Hamarneh G. BrainNetCNN: convolutional neural networks for brain networks; towards predicting neurodevelopment. Neuroimage. 2017;146:1038-1049.

    Not in our library10.1016/jneuroimage.2016.09.046

31. 31.

    Kingma DP, Ba J. Adam: a method for stochastic optimization. arXiv preprint arXiv: 1412.6980. 2014.

    Not in our library

32. 32.

    Kohl SAA, Romera-Paredes B, Meyer C, De Fauw J, Ledsam JR, Maier-Hein KH, Ali Eslami SM, Rezende DJ, Ronneberger O. A probabilistic U-net for segmentation of ambiguous images. 2018;6965-6975.

    Not in our library

33. 33.

    Kolbeinsson A, Kossaifi J, Panagakis Y, Bulat A, Anandkumar A, Tzoulaki I, Matthews P. Robust deep networks with randomized tensor regression layers arXiv Prepr. arXiv1902.10758. 2019.

    Not in our library

34. 34.

    Korolev S, Safiullin A, Belyaev M, Dodonova Y. Residual and plain convolutional neural networks for 3D brain MRI classification. 2017;835-838.

    Not in our library10.1109/isbi.2017.7950647

35. 35.

    Krizhevsky A, Sutskever I, Hinton GE. ImageNet classification with deep convolutional neural networks. Advances in Neural Information Processing Systems. 2012;1097-1105.

    Not in our library

36. 36.
    Deep learning · 2015

    No notice
37. 37.
    Convolutional Neural Networks-Based MRI Image Analysis for the Alzheimer’s Disease Prediction From Mild Cognitive Impairment · 2018

    No notice
38. 38.

    Liu M, Zhang J, Adeli E, Shen D. Landmark-based deep multi-instance learning for brain disease diagnosis. Med Image Anal. 2018;43:157-168.

    Not in our library10.1016/jmedia.2017.10.005

39. 39.

    Long J, Shelhamer E, Darrell T. Fully convolutional networks for semantic segmentation. 2015;3431-3440.

    Not in our library10.1109/tpami.2016.2572683

40. 40.

    MacMahon S, Peto R, Collins R, Godwin J, MacMahon S, Cutler J, Sorlie P, Abbott R, Collins R, Neaton J, Abbott R, et al. Blood pressure, stroke, and coronary heart disease: part 1, prolonged differences in blood pressure: prospective observational studies corrected for the regression dilution bias. Lancet. 1990;335:765-774.

    Not in our library10.1016/0140-6736(90)90878-9

Cited by papers we hold 5

• Mind the gap: Performance metric evaluation in brain‐age prediction · 2022
  No notice
• Individual variation underlying brain age estimates in typical development · 2021
  Not checked
• Deep learning encodes robust discriminative neuroimaging representations to outperform standard machine learning · 2021
  No notice
• Classifying Autism Spectrum Disorder Using the Temporal Statistics of Resting-State Functional MRI Data With 3D Convolutional Neural Networks · 2020
  Not checked
• Different scaling of linear models and deep learning in UKBiobank brain images versus machine-learning datasets · 2020
  No notice

Reads

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.