A state-space model of the hemodynamic approach: nonlinear filtering of BOLD signals

In this paper, a new procedure is presented which allows the estimation of the states and parameters of the hemodynamic approach from blood oxygenation level dependent (BOLD) responses. The proposed method constitutes an alternative to the recently proposed Friston [Neuroimage 16 (2002) 513] method and has some advantages over it. The procedure is based on recent groundbreaking time series analysis techniques that have been, in this case, adopted to characterize hemodynamic responses in functional magnetic resonance imaging (fMRI). This work represents a fundamental improvement over existing approaches to system identification using nonlinear hemodynamic models and is important for three reasons. First, our model includes physiological noise. Previous models have been based upon ordinary differential equations that only allow for noise or error to enter at the level of observation. Secondly, by using the innovation method and the local linearization filter, not only the parameters, but also the underlying states of the system generating responses can be estimated. These states can include things like a flow-inducing signal triggered by neuronal activation, de-oxyhemoglobine, cerebral blood flow and volume. Finally, radial basis functions have been introduced as a parametric model to represent arbitrary temporal input sequences in the hemodynamic approach, which could be essential to understanding those brain areas indirectly related to the stimulus. Hence, thirdly, by inferring about the radial basis parameters, we are able to perform a blind deconvolution, which permits both the reconstruction of the dynamics of the most likely hemodynamic states and also, to implicitly reconstruct the underlying synaptic dynamics, induced experimentally, which caused these states variations. From this study, we conclude that in spite of the utility of the standard discrete convolution approach used in statistical parametric maps (SPM), nonlinear BOLD phenomena and unspecific input temporal sequences must be included in the fMRI analysis.

[1]  Scott A. Huettel,et al.  Regional Differences in the Refractory Period of the Hemodynamic Response: An Event-Related fMRI Study , 2001, NeuroImage.

[2]  J. R. Baker,et al.  The intravascular contribution to fmri signal change: monte carlo modeling and diffusion‐weighted studies in vivo , 1995, Magnetic resonance in medicine.

[3]  T. Sejnowski,et al.  Single-Trial Variability in Event-Related BOLD Signals , 2002, NeuroImage.

[4]  M. Raichle,et al.  The Effects of Changes in PaCO2 Cerebral Blood Volume, Blood Flow, and Vascular Mean Transit Time , 1974, Stroke.

[5]  Tohru Ozaki.  Identification of nonlinearities and non-Gaussianities in time series , 1992 .

[6]  Juan C. Jiménez,et al.  Linear estimation of continuous-discrete linear state space models with multiplicative noise , 2002, Syst. Control. Lett..

[7]  B. Rosen,et al.  Evidence of a Cerebrovascular Postarteriole Windkessel with Delayed Compliance , 1999, Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism.

[8]  Robert Costalat,et al.  A Model of the Coupling between Brain Electrical Activity, Metabolism, and Hemodynamics: Application to the Interpretation of Functional Neuroimaging , 2002, NeuroImage.

[9]  Tohru Ozaki,et al.  An Approximate Innovation Method For The Estimation Of Diffusion Processes From Discrete Data , 2006 .

[10]  A. Dale,et al.  Functional–Anatomic Study of Episodic Retrieval II. Selective Averaging of Event-Related fMRI Trials to Test the Retrieval Success Hypothesis , 1998, NeuroImage.

[11]  C. Loan.  Computing integrals involving the matrix exponential , 1978 .

[12]  T. Ozaki.  2 Non-linear time series models and dynamical systems , 1985 .

[13]  Eric R. Ziecel.  Proceedings of the First U.S./Japan Conference on the Frontiers of Statistical Modeling: An Informational Approach , 1994 .

[14]  R. Buxton,et al.  Dynamics of blood flow and oxygenation changes during brain activation: The balloon model , 1998, Magnetic resonance in medicine.

[15]  J. C. Jimenez,et al.  A numerical method for the computation of the Lyapunov exponents of nonlinear ordinary differential equations , 2002, Appl. Math. Comput..

[16]  Karl J. Friston,et al.  Bayesian Estimation of Dynamical Systems: An Application to fMRI , 2002, NeuroImage.

[17]  P. Bandettini,et al.  Spatial Heterogeneity of the Nonlinear Dynamics in the FMRI BOLD Response , 2001, NeuroImage.

[18]  R. C. Oldfield.  The assessment and analysis of handedness: the Edinburgh inventory. , 1971, Neuropsychologia.

[19]  M. D’Esposito,et al.  The Variability of Human, BOLD Hemodynamic Responses , 1998, NeuroImage.

[20]  Ravi S. Menon,et al.  Functional brain mapping by blood oxygenation level-dependent contrast magnetic resonance imaging. A comparison of signal characteristics with a biophysical model. , 1993, Biophysical journal.

[21]  A M Dale,et al.  Estimation and detection of event‐related fMRI signals with temporally correlated noise: A statistically efficient and unbiased approach , 2000, Human brain mapping.

[22]  G. McCarthy,et al.  Evidence for a Refractory Period in the Hemodynamic Response to Visual Stimuli as Measured by MRI , 2000, NeuroImage.

[23]  Juan Carlos Jimenez,et al.  A simple algebraic expression to evaluate the local linearization schemes for stochastic differential equations , 2002, Appl. Math. Lett..

[24]  R. J. Biscay,et al.  Approximation of continuous time stochastic processes by the local linearization method revisited , 2002 .

[25]  J. C. Jimenez,et al.  Local Linearization method for the numerical solution of stochastic differential equations , 1996 .

[26]  P. Magistretti,et al.  Cellular mechanisms of brain energy metabolism and their relevance to functional brain imaging. , 1999, Philosophical transactions of the Royal Society of London. Series B, Biological sciences.

[27]  Karl J. Friston,et al.  Nonlinear Responses in fMRI: The Balloon Model, Volterra Kernels, and Other Hemodynamics , 2000, NeuroImage.

[28]  Tohru Ozaki.  The Local Linearization Filter with Application to Nonlinear System Identifications , 1994 .

[29]  C. Iadecola,et al.  Intrinsic signals and functional brain mapping: caution, blood vessels at work. , 2002, Cerebral cortex.

[30]  J. C. Jimenez,et al.  Simulation of Stochastic Differential Equations Through the Local Linearization Method. A Comparative Study , 1999 .

[31]  Juan C. Jiménez,et al.  Nonlinear EEG analysis based on a neural mass model , 1999, Biological Cybernetics.

[32]  J. C. Jimenez,et al.  Local linearization filters for non-linear continuous-discrete state space models with multiplicative noise , 2003 .

[33]  Rolando J. Biscay,et al.  Computing the noise covariance matrix of the local linearization scheme for the numerical solution of stochastic differential equations , 1998 .

[34]  Lars Kai Hansen,et al.  Modeling the hemodynamic response in fMRI using smooth FIR filters , 2000, IEEE Transactions on Medical Imaging.

[35]  Tohru Ozaki,et al.  The Role of the Likelihood Function in the Estimation of Chaos Models , 2000 .

[36]  Juan C. Jiménez,et al.  Dynamic properties of the local linearization method for initial-value problems , 2002, Appl. Math. Comput..

[37]  R. Cox,et al.  Event‐related fMRI contrast when using constant interstimulus interval: Theory and experiment , 2000, Magnetic resonance in medicine.

[38]  H. Benali,et al.  Robust Bayesian estimation of the hemodynamic response function in event‐related BOLD fMRI using basic physiological information , 2003, Human brain mapping.

[39]  Ying Zheng,et al.  A Model of the Hemodynamic Response and Oxygen Delivery to Brain , 2002, NeuroImage.

[40]  J. Rajapakse,et al.  Human Brain Mapping 6:283–300(1998) � Modeling Hemodynamic Response for Analysis of Functional MRI Time-Series , 2022 .

[41]  T. Ozaki.  A local linearization approach to nonlinear filtering , 1993 .

[42]  Karl J. Friston,et al.  Nonlinear Coupling between Evoked rCBF and BOLD Signals: A Simulation Study of Hemodynamic Responses , 2001, NeuroImage.

[43]  Karl J. Friston,et al.  Analysis of functional MRI time‐series , 1994, Human Brain Mapping.

[44]  J. Runnenburg.  PROBABILITY THEORY AND ITS APPLICATIONS , 1985 .

[45]  Scott L. Zeger,et al.  Non‐linear Fourier Time Series Analysis for Human Brain Mapping by Functional Magnetic Resonance Imaging , 1997 .

[46]  Héctor de Arazoza,et al.  Nonlinear Parametric Model Identification Using Genetic Algorithms , 2003, IWANN.

[47]  Angela R Laird,et al.  Characterizing instantaneous phase relationships in whole‐brain fMRI activation data , 2002, Human brain mapping.

[48]  Karl J. Friston,et al.  Nonlinear event‐related responses in fMRI , 1998, Magnetic resonance in medicine.

[49]  G. Berns,et al.  Continuous Functional Magnetic Resonance Imaging Reveals Dynamic Nonlinearities of “Dose-Response” Curves for Finger Opposition , 1999, The Journal of Neuroscience.