Interpretable brain age prediction using linear latent variable models of functional connectivity

Neuroimaging-driven prediction of brain age, defined as the predicted biological age of a subject using only brain imaging data, is an exciting avenue of research. In this work we seek to build models of brain age based on functional connectivity while prioritizing model interpretability and understanding. This way, the models serve to both provide accurate estimates of brain age as well as allow us to investigate changes in functional connectivity which occur during the ageing process. The methods proposed in this work consist of a two-step procedure: first, linear latent variable models, such as PCA and its extensions, are employed to learn reproducible functional connectivity networks present across a cohort of subjects. The activity within each network is subsequently employed as a feature in a linear regression model to predict brain age. The proposed framework is employed on the data from the CamCAN repository and the inferred brain age models are further demonstrated to generalize using data from two open-access repositories: the Human Connectome Project and the ATR Wide-Age-Range.

[1]  Stuart J. Ritchie,et al.  Brain age predicts mortality , 2017, Molecular Psychiatry.

[2]  Robert Leech,et al.  Efficiently searching through large tACS parameter spaces using closed-loop Bayesian optimization , 2019, Brain Stimulation.

[3]  Tor D Wager,et al.  The relation between statistical power and inference in fMRI , 2017, PloS one.

[4]  Takeshi Ogawa,et al.  Large-scale brain network associated with creative insight: combined voxel-based morphometry and resting-state functional connectivity analyses , 2018, Scientific Reports.

[5]  N. Maurits,et al.  A Brain-Wide Study of Age-Related Changes in Functional Connectivity. , 2015, Cerebral cortex.

[6]  Essa Yacoub,et al.  The WU-Minn Human Connectome Project: An overview , 2013, NeuroImage.

[7]  C. Grady,et al.  Age differences in the functional interactions among the default, frontoparietal control, and dorsal attention networks , 2016, Neurobiology of Aging.

[8]  Jianyi Lin,et al.  Short-Term Classification Learning Promotes Rapid Global Improvements of Information Processing in Human Brain Functional Connectome , 2020, Frontiers in Human Neuroscience.

[9]  M. Just,et al.  Functional connectivity in a baseline resting-state network in autism , 2006, Neuroreport.

[10]  Nasser M. Nasrabadi,et al.  Pattern Recognition and Machine Learning , 2006, Technometrics.

[11]  Karl J. Friston,et al.  A Voxel-Based Morphometric Study of Ageing in 465 Normal Adult Human Brains , 2001, NeuroImage.

[12]  Christos Davatzikos,et al.  Accelerated brain aging in schizophrenia and beyond: a neuroanatomical marker of psychiatric disorders. , 2014, Schizophrenia bulletin.

[13]  Christian Gaser,et al.  Advanced BrainAGE in older adults with type 2 diabetes mellitus , 2013, Front. Aging Neurosci..

[14]  Dimitri P. Bertsekas,et al.  Constrained Optimization and Lagrange Multiplier Methods , 1982 .

[15]  Michael B. Miller,et al.  How reliable are the results from functional magnetic resonance imaging? , 2010, Annals of the New York Academy of Sciences.

[16]  N. Raz,et al.  Differential Aging of the Brain: Patterns, Cognitive Correlates and Modifiers , 2022 .

[17]  R. Henson,et al.  Challenges in measuring individual differences in functional connectivity using fMRI: The case of healthy aging , 2017, Human brain mapping.

[18]  Christoforos Anagnostopoulos,et al.  Decoding Time-Varying Functional Connectivity Networks via Linear Graph Embedding Methods , 2017, Front. Comput. Neurosci..

[19]  Ricardo Pio Monti,et al.  Dissociating frontoparietal brain networks with neuroadaptive Bayesian optimization , 2017, Nature Communications.

[20]  Aapo Hyvärinen,et al.  Variational Autoencoders and Nonlinear ICA: A Unifying Framework , 2019, AISTATS.

[21]  Robert Leech,et al.  Prediction of brain age suggests accelerated atrophy after traumatic brain injury , 2015, Annals of neurology.

[22]  Jonathan D. Power,et al.  Prediction of Individual Brain Maturity Using fMRI , 2010, Science.

[23]  Stephen M. Smith,et al.  The future of FMRI connectivity , 2012, NeuroImage.

[24]  Kuncheng Li,et al.  Changes of functional connectivity of the motor network in the resting state in Parkinson's disease , 2009, Neuroscience Letters.

[25]  Katherine E. Prater,et al.  Functional connectivity tracks clinical deterioration in Alzheimer's disease , 2012, Neurobiology of Aging.

[26]  Carey E. Priebe,et al.  Statistical Inference on Random Dot Product Graphs: a Survey , 2017, J. Mach. Learn. Res..

[27]  Joachim M. Buhmann,et al.  Expectation-maximization for sparse and non-negative PCA , 2008, ICML '08.

[28]  Christoforos Anagnostopoulos,et al.  Estimating time-varying brain connectivity networks from functional MRI time series , 2013, NeuroImage.

[29]  Giovanni Montana,et al.  Predicting brain age with deep learning from raw imaging data results in a reliable and heritable biomarker , 2016, NeuroImage.

[30]  Aapo Hyvärinen,et al.  Independent component analysis of fMRI group studies by self-organizing clustering , 2005, NeuroImage.

[31]  Aapo Hyvärinen,et al.  Fast and robust fixed-point algorithms for independent component analysis , 1999, IEEE Trans. Neural Networks.

[32]  Gary H. Glover,et al.  Reducing interscanner variability of activation in a multicenter fMRI study: Controlling for signal-to-fluctuation-noise-ratio (SFNR) differences , 2006, NeuroImage.

[33]  E. B. Andersen,et al.  Modern factor analysis , 1961 .

[34]  Aapo Hyvärinen,et al.  A unified probabilistic model for learning latent factors and their connectivities from high-dimensional data , 2018, UAI.

[35]  Christoforos Anagnostopoulos,et al.  Real‐time estimation of dynamic functional connectivity networks , 2017, Human brain mapping.

[36]  Mark W. Woolrich,et al.  Advances in functional and structural MR image analysis and implementation as FSL , 2004, NeuroImage.

[37]  J. Cole,et al.  Predicting Age Using Neuroimaging: Innovative Brain Ageing Biomarkers , 2017, Trends in Neurosciences.

[38]  Motoaki Kawanabe,et al.  Orthogonal Connectivity Factorization: Interpretable Decomposition of Variability in Correlation Matrices , 2016, Neural Computation.

[39]  O. Sporns Discovering the Human Connectome , 2012 .

[40]  Aapo Hyvärinen,et al.  Validating the independent components of neuroimaging time series via clustering and visualization , 2004, NeuroImage.

[41]  B. Biswal,et al.  Characterizing variation in the functional connectome: promise and pitfalls , 2012, Trends in Cognitive Sciences.

[42]  Thomas E. Nichols,et al.  Handbook of Functional MRI Data Analysis: Index , 2011 .

[43]  Thomas E. Nichols,et al.  Estimation of Brain Age Delta from Brain Imaging , 2019 .

[44]  Ricardo Pio Monti,et al.  Classifying HCP task-fMRI networks using heat kernels , 2016, 2016 International Workshop on Pattern Recognition in Neuroimaging (PRNI).

[45]  V. Calhoun,et al.  The Chronnectome: Time-Varying Connectivity Networks as the Next Frontier in fMRI Data Discovery , 2014, Neuron.

[46]  Russell A. Poldrack,et al.  Handbook of Functional MRI Data Analysis: Visualizing, localizing, and reporting fMRI data , 2011 .

[47]  Michael Eickenberg,et al.  Machine learning for neuroimaging with scikit-learn , 2014, Front. Neuroinform..

[48]  Dimitri Van De Ville,et al.  Principal components of functional connectivity: A new approach to study dynamic brain connectivity during rest , 2013, NeuroImage.

[49]  Ludovica Griffanti,et al.  Automatic denoising of functional MRI data: Combining independent component analysis and hierarchical fusion of classifiers , 2014, NeuroImage.

[50]  H. Hotelling Analysis of a complex of statistical variables into principal components. , 1933 .

[51]  Timothy O. Laumann,et al.  Functional Network Organization of the Human Brain , 2011, Neuron.

[52]  Jenessa Lancaster,et al.  Bayesian Optimization for Neuroimaging Pre-processing in Brain Age Classification and Prediction , 2018, Front. Aging Neurosci..

[53]  E. Formisano,et al.  Functional connectivity as revealed by spatial independent component analysis of fMRI measurements during rest , 2004, Human brain mapping.

[54]  John Ashburner,et al.  Computational anatomy with the SPM software. , 2009, Magnetic resonance imaging.

[55]  P. Uhlhaas,et al.  Preferential Detachment During Human Brain Development: Age- and Sex-Specific Structural Connectivity in Diffusion Tensor Imaging (DTI) Data , 2013, Cerebral cortex.

[56]  I. Jolliffe Principal Component Analysis , 2002 .

[57]  Motoaki Kawanabe,et al.  Characterizing Variability of Modular Brain Connectivity with Constrained Principal Component Analysis , 2016, PloS one.

[58]  Cam-CAN Group,et al.  The Cambridge Centre for Ageing and Neuroscience (Cam-CAN) data repository: Structural and functional MRI, MEG, and cognitive data from a cross-sectional adult lifespan sample , 2017, NeuroImage.

[59]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[60]  Christoforos Anagnostopoulos,et al.  Learning population and subject-specific brain connectivity networks via Mixed Neighborhood Selection , 2015, 1512.01947.

[61]  Christoforos Anagnostopoulos,et al.  Graph Embeddings of Dynamic Functional Connectivity Reveal Discriminative Patterns of Task Engagement in HCP Data , 2015, 2015 International Workshop on Pattern Recognition in NeuroImaging.

[62]  Stefan Klöppel,et al.  Estimating the age of healthy subjects from T1-weighted MRI scans using kernel methods: Exploring the influence of various parameters , 2010, NeuroImage.

[63]  Linda Geerligs,et al.  Reduced specificity of functional connectivity in the aging brain during task performance , 2014, Human brain mapping.

[64]  Linda Geerligs,et al.  State and Trait Components of Functional Connectivity: Individual Differences Vary with Mental State , 2015, The Journal of Neuroscience.

[65]  P. Dayan,et al.  Early childhood investment impacts social decision-making four decades later , 2018, Nature Communications.