Pharmacopsychiatry 2007; 40: S85-S92
DOI: 10.1055/s-2007-992769
Original Paper

© Georg Thieme Verlag KG Stuttgart · New York

Assessing the State Space of the Brain with fMRI: An Integrative View of Current Methods

R. G. M. Schlösser 1 , K. Koch 1 , G. Wagner 1
  • 1Department of Psychiatry and Psychotherapy, University of Jena, Jena, Germany
Further Information

Publication History

Publication Date:
17 December 2007 (online)

Abstract

Systems biology has gained substantial benefit from the application of systems modeling in engineering sciences. In general, methods as employed for construction and simulation of technical devices and buildings are applicable to modeling of biological systems. A number of modeling approaches originally derived from different areas such as engineering, econometrics and genetics have been adapted to functional brain imaging datasets in the recent years. However, despite a number of analogies, the complexities of brain systems might be much higher than those observed in technical systems. A dynamical system can be described as a state space in which a certain state of the system is specified by a single point. Varying states of the system over time can be described by a trajectory of states. Different modeling algorithms focus on certain aspects of this state space. The covariance of the state-space variables can be examined by correlational analysis (targeting normalized covariance) and principal component analysis. One of the principle aims in any systems identification approach is to identify parameters of the state matrix, i.e. the rules of transitions between different states of the system. Dynamic approaches with temporal information include the full state space model, vector autoregressive model and dynamic causal modeling. Structural equation modeling focuses on the instantaneous relationship between functional nodes. Directional analysis strategies are available in temporal and frequency domain. Depending on general assumptions as to how neuronal representation is established, the approaches present complementary information about the underlying neuronal interactions. The present article attempts to provide an integrative overview of the most established models and methods which are currently being applied for modeling dynamic brain systems.

References

  • 1 Arnold M, Miltner WH, Witte H, Bauer R, Braun C. Adaptive AR modeling of nonstationary time series by means of Kalman filtering.  IEEE transactions on bio-medical engineering. 1998;  45 553-562
  • 2 Astolfi L, Cincotti F, Mattia D. et al . Estimation of the effective and functional human cortical connectivity with structural equation modeling and directed transfer function applied to high-resolution EEG.  Magn Reson Imaging. 2004;  22 1457-1470
  • 3 Aström KJ, Wittenmark B. Adaptive control. Rading: Addison-Wesley Publishing Company 1995
  • 4 Basar-Eroglu C, Brand A, Hildebrandt H. et al . Working memory related gamma oscillations in schizophrenia patients.  Int J Psychophysiol. 2007;  64 39-45
  • 5 Bender W, Albus M, Moller HJ, Tretter F. Towards systemic theories in biological psychiatry.  Pharmacopsychiatry. 2006;  39 (Suppl 1) S4-S9
  • 6 Bhattacharya S, Ringo Ho MH, Purkayastha S. A Bayesian approach to modeling dynamic effective connectivity with fMRI data.  Neuroimage. 2006;  30 794-812
  • 7 Boomsma A, Hoogland JJ. The robustness of LISREL modeling revisited. In: Cudeck R, du Toit S, Sörbom D eds, Structural Equation Modeling: Present and Future. Lincolnwood, IL: Scientific Software International 2001: 139-168
  • 8 Brogan WL. Modern control theory. 3 ed. Upper Saddle River. NJ: Prentice Hall 1991
  • 9 Buchel C, Friston KJ. Modulation of connectivity in visual pathways by attention: cortical interactions evaluated with structural equation modelling and fMRI.  Cereb Cortex. 1997;  7 768-778
  • 10 Buchel C, Friston KJ. Dynamic changes in effective connectivity characterized by variable parameter regression and Kalman filtering.  Hum Brain Mapp. 1998;  6 403-408
  • 11 Bullmore E, Horwitz B, Honey G. et al . How good is good enough in path analysis of fMRI data?.  Neuroimage. 2000;  11 289-301
  • 12 Byrne BM. Structural equation modeling with AMOS. Basic concepts, applications, and programming. Mahwah, New Jersey: Lawrence Erlbaum Associates 2001
  • 13 Carmeli C, Knyazeva MG, Innocenti GM, Feo O De. Assessment of EEG synchronization based on state-space analysis.  Neuroimage. 2005;  25 339-354
  • 14 Chatfield C. The analysis of time series: An introduction. Boca Raton: Chapman & Hall/CRC 2004
  • 15 Cordes D, Haughton VM, Arfanakis K. et al . Mapping functionally related regions of brain with functional connectivity MR imaging.  AJNR Am J Neuroradiol. 2000;  21 1636-1644
  • 16 Deco G. A dynamical model of event-related FMRI signals in prefrontal cortex: predictions for schizophrenia.  Pharmacopsychiatry. 2006;  39 (Suppl 1) S65-S67
  • 17 Dempster AP, Laird NM. Maximum likelihood estimation from incomplete data via the EM algorithm.  J Roy Statist Soc B. 1977;  39 1-38
  • 18 Doya K, Ishii S, Pouget A. Bayesian Brain: Probabilistic Approaches to Neural Coding MIT Press 2007
  • 19 Eliasmith C, Anderson CH. Neural Engineering. Cambridge, Massachusetts: MIT Press 2003
  • 20 Etkin A, Egner T, Peraza DM, Kandel ER, Hirsch J. Resolving emotional conflict: A role for the rostral anterior cingulate cortex in modulating activity in the amygdala.  Neuron. 2006;  51 871-882
  • 21 Fell J, Fernandez G, Klaver P, Elger CE, Fries P. Is synchronized neuronal gamma activity relevant for selective attention?.  Brain Res Brain Res Rev. 2003;  42 265-272
  • 22 Friman O, Farneback G, Westin CF. A Bayesian approach for stochastic white matter tractography.  IEEE transactions on medical imaging. 2006;  25 965-978
  • 23 Friston K. A theory of cortical responses.  Philosophical transactions of the Royal Society of London. 2005;  360 815-836
  • 24 Friston KJ, Harrison L, Penny W. Dynamic causal modelling.  Neuroimage. 2003;  19 1273-1302
  • 25 Gallinat J, Heinz A. Combination of multimodal imaging and molecular genetic information to investigate complex psychiatric disorders.  Pharmacopsychiatry. 2006;  39 (Suppl 1) S76-S79
  • 26 Gelman A, Carlin JB, Stern HS, Rubin DB. Bayesian data analysis. Boca Raton: Cahpman & Hall 2003
  • 27 Gerstner W. Kistler w, Spiking Neuron Models: Single Neurons, Populations, Plasticity. Cambridge: Cambridge University Press 2002
  • 28 Geweke J. Measurement of linear dependence and feedback between multiple time series.  Journal of the American Statistical Association. 1982;  77 304-313
  • 29 Goebel R, Roebroeck A, Kim DS, Formisano E. Investigating directed cortical interactions in time-resolved fMRI data using vector autoregressive modeling and Granger causality mapping.  Magn Reson Imaging. 2003;  21 1251-1261
  • 30 Gossl C, Fahrmeir L, Putz B, Auer LM, Auer DP. Fiber tracking from DTI using linear state space models: detectability of the pyramidal tract.  Neuroimage. 2002;  16 378-388
  • 31 Granger CWJ. Investigating causal relations by econometric models and cross-spectral methods.  Econometrica. 1969;  37 424-438
  • 32 Grewal MS, Andrews AP. Kalman Filtering: Theory and practice using MATLAB: Wiley-Interscience 2001
  • 33 Haavisto O, Hyotyniemi H, Roos C. State space modeling of yeast gene expression dynamics.  Journal of bioinformatics and computational biology. 2007;  5 31-46
  • 34 Harrison L, Penny WD, Friston K. Multivariate autoregressive modeling of fMRI time series.  Neuroimage. 2003;  19 1477-1491
  • 35 Ho RM, Ombao H, Shumway R. A state-space approach to modeling brain dynamics.  Statistica Sinica. 2005;  15 407-425
  • 36 Honey G, Fu C, Kim J. et al . Effects of verbal working memory load on corticocortical connectivity modeled by path analysis of functional magnetic resonance imaging data.  Neuroimage. 2002;  17 573-582
  • 37 Honey GD, Suckling J, Zelaya F. et al . Dopaminergic drug effects on physiological connectivity in a human cortico-striato-thalamic system.  Brain. 2003;  126 1767-1781
  • 38 Horwitz B, MacIntosh AR, Haxby JV, Grady CL. Network analysis of brain cognitive function using metabolic and blood flow data.  Behav Brain Res. 1995;  66 187-193
  • 39 Hyvärinen A, Karhunen J, Oja E. Independent component analysis. New York: John Wiley & Sons, Inc 2001
  • 40 Kalman RE. A new approach to linear filtering and prediction problems.  Trans of the ASME - Journal of Basic Engineering. 1960;  35-45
  • 41 Kaminski M, Ding M, Truccolo WA, Bressler SL. Evaluating causal relations in neural systems: granger causality, directed transfer function and statistical assessment of significance.  Biological cybernetics. 2001;  85 145-157
  • 42 Kim J, Zhu W, Chang L, Bentler PM, Ernst T. Unified structural equation modeling approach for the analysis of multisubject, multivariate functional MRI data.  Hum Brain Mapp. 2006; 
  • 43 Kline RB. Principles and practice of structural equation modeling. 2nd ed. New York: The Guilford Press 2005
  • 44 Kobayashi R, Shinomoto S. State space method for predicting the spike times of a neuron.  Physical review. 2007;  75 011925
  • 45 Koch C, Hepp K. Quantum mechanics in the brain.  Nature. 2006;  440 611
  • 46 Koechlin E, Ody C, Kouneiher F. The architecture of cognitive control in the human prefrontal cortex.  Science. 2003;  302 1181-1185
  • 47 Lathi BP. Linear Systems and Signals. 2 ed. Oxford: Oxford University Press 2005
  • 48 Ljung L. System identification: Theory for the User. 2nd edition ed. London: Prentice Hall 1999
  • 49 MacArdle J. Causal modeling applied to psychonomic systems simulation.  Behavior Research Methods and Instrumentation. 1980;  12 193-209
  • 50 MacIntosh AR, Bookstein FL, Haxby JV, Grady CL. Spatial pattern analysis of functional brain images using partial least squares.  Neuroimage. 1996;  3 143-157
  • 51 MacIntosh AR, Grady CL, Ungerleider LG. et al . Network analysis of cortical visual pathways mapped with PET.  J Neurosci. 1994;  14 655-666
  • 52 Mechelli A, Price CJ, Noppeney U, Friston KJ. A dynamic causal modeling study on category effects: bottom-up or top-down mediation?.  J Cogn Neurosci. 2003;  15 925-934
  • 53 Musso F, Konrad A, Vucurevic G. et al . Distributed BOLD-response in association cortex vector state space predicts reaction time during selective attention.  Neuroimage. 2006;  29 1311-1318
  • 54 Noppeney U, Josephs O, Hocking J, Price CJ, Friston KJ. The effect of prior visual information on recognition of speech and sounds.  Cereb Cortex. 2007; 
  • 55 Paulus MP, Frank L, Brown GG, Braff DL. Schizophrenia subjects show intact success-related neural activation but impaired uncertainty processing during decision-making.  Neuropsychopharmacology. 2003;  28 795-806
  • 56 Penny WD, Stephan KE, Mechelli A, Friston KJ. Modelling functional integration: a comparison of structural equation and dynamic causal models.  Neuroimage. 2004;  23 (Suppl 1) S264-S274
  • 57 Petrides M. Mapping prefrontal cortical systems for the control of cognition. In: Toga A, Mazziotta JC eds, Brain mapping: the systems. San Diego: Academic Press 2000: 157-176
  • 58 Posner MI. Chronometric Explorations of Mind. Hillsdale, N.J.: Lawrence Erlbaum Associates 1978
  • 59 Rao RPN. Neural models of Bayesian belief propagation. In: Doya K, Ishii S, Pouget A, Rao RPN eds, Bayesian Brain: Probabilistic Approaches to Neural Coding Cambridge. MA: MIT Press 2007: 235-264
  • 60 Roebroeck A, Formisano E, Goebel R. Mapping directed influence over the brain using Granger causality and fMRI.  Neuroimage. 2005;  25 230-242
  • 61 Sato JR, Junior EA, Takahashi DY. et al . A method to produce evolving functional connectivity maps during the course of an fMRI experiment using wavelet-based time-varying Granger causality.  Neuroimage. 2006; 
  • 62 Schild D. Coordination of neuronal signals as structures in state space.  The International journal of neuroscience. 1984;  22 283-297
  • 63 Schlösser R, Gesierich T, Kaufmann B. et al . Altered effective connectivity during working memory performance in schizophrenia: a study with fMRI and structural equation modeling.  Neuroimage. 2003;  19 751-763
  • 64 Schlösser R, Gesierich T, Kaufmann B, Vucurevic G, Stoeter P. Altered effective connectivity in drug free schizophrenic patients.  Neuroreport. 2003;  14 2233-2237
  • 65 Steele JD, Meyer M, Ebmeier KP. Neural predictive error signal correlates with depressive illness severity in a game paradigm.  Neuroimage. 2004;  23 269-280
  • 66 Summerfield C, Egner T, Greene M. et al . Predictive codes for forthcoming perception in the frontal cortex.  Science. 2006;  314 1311-1314
  • 67 Toni I, Rowe J, Stephan KE, Passingham RE. Changes of cortico-striatal effective connectivity during visuomotor learning.  Cereb Cortex. 2002;  12 1040-1047
  • 68 Heijden F van der, Duin R, Ridder D de, Tax DMJ. Classification, Parameter Estimation and State Estimation: An Engineering Approach Using MATLAB. New York: John Wiley & Sons 2004
  • 69 Winterhalder M, Schelter B, Hesse W. et al . Detection of directed information flow in biosignals.  Biomedizinische Technik. 2006;  51 281-287
  • 70 Wright JJ, Kydd RR, Lees GJ. State-changes in the brain viewed as linear steady-states and non-linear transitions between steady-states.  Biological cybernetics. 1985;  53 11-17
  • 71 Zhan Y, Halliday D, Jiang P, Liu X, Feng J. Detecting time-dependent coherence between non-stationary electrophysiological signals-A combined statistical and time-frequency approach.  Journal of neuroscience methods. 2006; 

1 Supported by grants BMBF FKZ01ZZ0105 and IZKF, TMWFK B30701-015/-016

Correspondence

R. G. M. SchlösserMD 

Department of Psychiatry and Psychotherapy

University of Jena

Philosophenweg 3

07740 Jena

Germany

Phone: +49/3641/9 352 84

Fax: +49/3641/9 354 44

Email: Ralf.Schloesser@uni-jena.de