Beyond linearity in neuroimaging: Capturing nonlinear relationships with application to longitudinal studies

The ubiquitous adoption of linearity for quantitative explanatory variables in statistical modeling is likely attributable to its advantages of straightforward interpretation and computational feasibility. The linearity assumption may be a reasonable approximation especially when the variable is confined within a narrow range, but it can be problematic when the variable’s effect is non-monotonic or complex. Furthermore, visualization and model assessment of a linear fit are usually omitted because of challenges at the whole brain level in neuroimaging. By adopting a principle of learning from the data in the presence of uncertainty to resolve the problematic aspects of conventional polynomial fitting, we introduce a flexible and adaptive approach of multilevel smoothing splines (MSS) to capture any nonlinearity of a quantitative explanatory variable for population-level neuroimaging data analysis. With no prior knowledge regarding the underlying relationship other than a parsimonious assumption about the extent of smoothness (e.g., no sharp corners), we express the unknown relationship with a sufficient number of smoothing splines and use the data to adaptively determine the specifics of the nonlinearity. In addition to introducing the theoretical framework of MSS as an efficient approach with a counterbalance between flexibility and stability, we strive to (a) lay out the specific schemes for population-level nonlinear analyses that may involve task (e.g., contrasting conditions) and subject-grouping (e.g., patients vs controls) factors; (b) provide modeling accommodations to adaptively reveal, estimate and compare any nonlinear effects of an explanatory variable across the brain, or to more accurately account for the effects (including nonlinear effects) of a quantitative confound; (c) offer the associated program 3dMSS to the neuroimaging community for whole-brain voxel-wise analysis as part of the AFNI suite; and (d) demonstrate the modeling approach and visualization processes with a longitudinal dataset of structural MRI scans.

[1]  Paul A Taylor,et al.  An integrative Bayesian approach to matrix‐based analysis in neuroimaging , 2019, Human brain mapping.

[2]  A. Dale,et al.  Critical ages in the life course of the adult brain: nonlinear subcortical aging , 2013, Neurobiology of Aging.

[3]  Hans-Georg Müller,et al.  Functional Data Analysis , 2016 .

[4]  Jean Duchon,et al.  Splines minimizing rotation-invariant semi-norms in Sobolev spaces , 1976, Constructive Theory of Functions of Several Variables.

[5]  Eva H. Telzer,et al.  Methodological considerations for developmental longitudinal fMRI research , 2018, Developmental Cognitive Neuroscience.

[6]  Susan J. Lederman,et al.  Multisensory Activation of the Intraparietal Area When Classifying Grating Orientation: A Functional Magnetic Resonance Imaging Study , 2006, The Journal of Neuroscience.

[7]  Christos Davatzikos,et al.  Harmonization of large MRI datasets for the analysis of brain imaging patterns throughout the lifespan , 2019, NeuroImage.

[8]  Guido Imbens,et al.  Why High-Order Polynomials Should Not Be Used in Regression Discontinuity Designs , 2014 .

[9]  Paul M. Thompson,et al.  Submission PDF Mapping Cortical Brain Asymmetry in 17 , 141 Healthy Individuals Worldwide via the ENIGMA Consortium , 2018 .

[10]  R W Cox,et al.  AFNI: software for analysis and visualization of functional magnetic resonance neuroimages. , 1996, Computers and biomedical research, an international journal.

[11]  Gang Chen,et al.  Handling Multiplicity in Neuroimaging Through Bayesian Lenses with Multilevel Modeling , 2017, Neuroinformatics.

[12]  Jennifer C. Britton,et al.  Linear mixed-effects modeling approach to FMRI group analysis , 2013, NeuroImage.

[13]  Paul A. Taylor,et al.  Fighting or embracing multiplicity in neuroimaging? neighborhood leverage versus global calibration , 2019, NeuroImage.

[14]  Gerard R. Ridgway,et al.  Symmetric Diffeomorphic Modeling of Longitudinal Structural MRI , 2013, Front. Neurosci..

[15]  Alan Y. Chiang,et al.  Generalized Additive Models: An Introduction With R , 2007, Technometrics.

[16]  Gilbert Saporta,et al.  L'analyse des données , 1981 .

[17]  Michael D. Gregory,et al.  Williams syndrome hemideletion and LIMK1 variation both affect dorsal stream functional connectivity. , 2019, Brain : a journal of neurology.

[18]  Klaus P. Ebmeier,et al.  Meta-analysis of generalized additive models in neuroimaging studies , 2020, NeuroImage.

[19]  R Core Team,et al.  R: A language and environment for statistical computing. , 2014 .

[20]  R. Tibshirani,et al.  Generalized Additive Models , 1986 .

[21]  M. Catani,et al.  A lateralized brain network for visuospatial attention , 2011, Nature Neuroscience.

[22]  Carl E. Rasmussen,et al.  Gaussian processes for machine learning , 2005, Adaptive computation and machine learning.

[23]  Mark S. Cohen,et al.  Parametric Analysis of fMRI Data Using Linear Systems Methods , 1997, NeuroImage.

[24]  Karolinska Schizophrenia Mapping cortical brain asymmetry in 17,141 healthy individuals worldwide via the ENIGMA consortium , 2018 .

[25]  Martyn McFarquhar,et al.  Modeling Group-Level Repeated Measurements of Neuroimaging Data Using the Univariate General Linear Model , 2019, Front. Neurosci..

[26]  Vince D. Calhoun,et al.  Brain Development Includes Linear and Multiple Nonlinear Trajectories: A Cross-Sectional Resting-State Functional Magnetic Resonance Imaging Study , 2019, Brain Connect..

[27]  Ellen Leibenluft,et al.  Applications of multivariate modeling to neuroimaging group analysis: A comprehensive alternative to univariate general linear model , 2014, NeuroImage.

[28]  S. Wood Thin plate regression splines , 2003 .

[29]  Gang Chen,et al.  To pool or not to pool: Can we ignore cross-trial variability in FMRI? , 2021, NeuroImage.

[30]  Virgilio Gómez-Rubio,et al.  Generalized Additive Models: An Introduction with R (2nd Edition) , 2018 .

[31]  Gang Chen,et al.  Detecting the subtle shape differences in hemodynamic responses at the group level , 2015, Front. Neurosci..

[32]  Robert J Zatorre,et al.  Asymmetries of the planum temporale and Heschl's gyrus: relationship to language lateralization. , 2006, Brain : a journal of neurology.

[33]  X. Bay,et al.  Generalization of the Kimeldorf-Wahba correspondence for constrained interpolation , 2016, 1602.02714.