| || |
|Title:||HARMONICS DECOMPOSITION FOR EVALUATION OF THE DIRECTION SELECTIVITY AND RECEPTIVE FIELD SIZE OF OPTICAL BRAIN IMAGING MAPS. |
| || |
|Authors/Affiliations:||2 Matthieu Vanni*; 1 Jean Provost; 2 Martin Villeneuve; 1 Frederic Lesage; 2 Christian Casanova; |
1 Ecole Polytechnique de Montreal; 2 Universite de Montreal, QC, Canada
| || |
|Content:||Objectives: The extraction of stimulus responses from noise has been and is still a fundamental issue in optical brain imaging research. A significant progress has been recently made by combining different conditions by periodically changing the stimuli, allowing then a strong diminution of the recording time (Kalatsky and Stryker (2003)). This paradigm is based on the spectral decomposition of the signal to give information about amplitude and phase locked to the stimulus. In this study, we investigated the possibility to extract addition information from the Fourier space. |
Materials and Methods: Recordings were performed on primary visual cortex (area 17, 18 and 19) and area 21a on eight anesthetised cats. For each pixel, absorbance was recorded during 10 to 30 min for a continuous and periodic stimulation. Fourier transform was computed in the temporal dimension to extract magnitude at harmonics of the stimulus frequency.
Results: First, we used higher harmonics as an estimation of the response profile. Using a sweeping bar in the visual field, the width of the periodic response was a convolution between the hemodynamic response function (HRF) and the receptive field (RF) activation. A narrow response evoked by small RFs was associated with a distribution of the signal in higher harmonics; a positive correlation between the ratio of the 2sd and 1st harmonics and spatial frequency (which is inversely proportional to the RF size) was observed. Second, we concentrated on the modular organization (orientation and direction selectivity domains) of the visual cortex. We investigated the responses to a pattern moving continuously at successive directions of a fixed periodicity. For a pattern stimulating direction selectivity (a random dots kinematogram), responses were found at the 1st harmonic level. Conversely, for static pattern stimulating orientation (a static grating), responses were confined to the 2sd harmonic.
Conclusion: Depending of the stimulus used, information from 1st and 2sd harmonics represents a powerful tool to measure the average RF size and direction selectivity. The relevance of using 3rd and 4th harmonics to evaluate direction tuning bandwidth is currently examined. Supp: NSERC to CC.
| || |