We have built a simulation to model the intrinsic signal sources. It uses estimates of parameter values derived from the published data from the optical imaging, the functional brain neuroimaging literature and textbook physiology.The temporal dynamics of the intrinsic signals are modelled by gamma functions to give delays with the appropriate time constants. The model generates a time series representing the image intensity signal under different wavelength illumination. Perturbations that mimic systemic noise sources such as heartbeat, breathing and vasomotion can be included as desired. The simulation provides a test bed for the evaluation of the effects of different data capture regimes and methods data analysis and a working hypothesis of the neuro-hemo dynamics that can be used in statistical and model driven analysis of intrinsic image data. In this study we describe the simulation of an ocular dominance column mapping experiment using the methods of intrinsic imaging. We find that though accurate functional maps could be obtained using the methods of analysis in common usage in intrinsic imagery, the estimates of the resolution of the signal sources obtained did not accurately reflect the underlying parameterisation of the model. We ascribe this to the presence of the low frequency modulation of regional cerebral blood flow now known to be present in intrinsic image data, which introduces bias when usual data capture and analysis methodology is used.
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)