Abstract:
Further standardization in signal processing tools is needed in the area of functional near infrared spectroscopy (fNIRS) before it is recognized as a reliable neuroimaging modality. This thesis study attempts to present a comprehensive analysis of the feasibility of applying statistical inference methods to fNIRS signals. Using hierarchical linear models, both classical and Bayesian techniques are pursued and performances of different methods are presented on a comparative basis. The results obtained from a set of cognitive signals show that fNIRS can identify cognitive activity both at the subject and group levels. The analysis suggests that mixed or Bayesian hierarchical models are especially convenient for fNIRS signals. A related problem that is discussed in this thesis study is to relate the outcome of the statistical analysis with the underlying physiology. This problem is studied by putting constraints over the parameters to be estimated. Carrying the problem to a Bayesian framework, the constraints were turned into prior distributions and Gibbs sampling was used to infer from the posterior distributions. The results exhibit that in addition to preventing unlikely results to appear at the end of the analysis, using parameter constraints is also more efficient in revealing activations which are obscured by heavy noise. The last part of this thesis study departs from hypothesis-based statistical inference techniques and introduces the use of information-theoretic measures for fNIRS by particularly concentrating on neural complexity and functional clustering. It is demonstrated that this type of measures may capture organizational aspects of the brain which are hard to reveal with classical statistical inference techniques.|Keywords: Functional near infrared spectroscopy, Statistical inference, Bayesian statistics, General linear model, Constrained estimation, Complexity.
