Evaluation of receptive field size from higher harmonics in visuotopic mapping using continuous stimulation optical imaging

https://doi.org/10.1016/j.jneumeth.2010.03.013Get rights and content

Abstract

The extraction and detection of specific responses from a large amount of background noise has been the subject of a considerable body of research in brain functional imaging, and more specifically in optical intrinsic signal imaging. Recent work by Kalatsky and Stryker (2003) showed that by combining different conditions and using periodic stimuli, recording times can be reduced. Spectral decomposition is then used to provide amplitude and phase information locked to the stimulus. A drawback of the above method is that by focusing only on a single harmonic, response information is limited. The shape of the hemodynamic response function (HRF) and the temporal variations in the neural responses cannot be assessed. In this work it is argued that additional information about neural responses can be gathered by using higher harmonics. Moving bars were used to generate visuotopic maps on large portions of the cat visual cortex. Up to four simultaneously bars moving repetitively across the visual field at different frequencies were used to sample the HRF in the Fourier domain. The HRF profile obtained with continuous stimulation was spatially homogeneous throughout the cortex and similar to the HRF profile obtained using episodic stimulation. Furthermore, by modeling the optical response as a convolution between HRF and neuronal responses, the ratio of the second harmonic to the first provided an estimation of the receptive field size. This was further validated by measuring spatial frequency selectivity. Therefore, the use of higher harmonics opens new avenues to estimate receptive field size from temporal signals.

Introduction

The extraction and detection of specific visual responses from a large amount of background noise remains a challenge in intrinsic optical imaging. In the literature, averaging or differential imaging methods (used in episodic paradigm) are generally used (Blasdel, 1992) and are probably the simplest means to extract neuronal maps. Averaging a number of repetitions exploits the fact that sources of physiological periodic noise (i.e., heartbeat, respiration and vasomotion) are not synchronized with the stimulation (Mayhew et al., 1996). Refinements on these analyses such as signal decomposition by principal components analysis have been subsequently developed (Everson et al., 1997, Gabbay et al., 2000). More recently, time-locked or periodic techniques which infer responses from amplitude and phase information locked to the stimulus have been explored (Sornborger et al., 2003, Sornborger et al., 2005). For example, using these principles, Kalatsky and Stryker (2003) revealed precise visuotopic maps of the visual cortex. In contrast to the episodic technique, the continuous stimulation paradigm allows the combination of different conditions by periodically changing the stimuli recovering the spatial response by using phase information at the stimulation frequency. Experiment duration is thus considerably reduced since, as in case of visuotopic studies, an “infinite” number of stimulus positions can be presented with a high signal to noise ratio (SNR). However, the detection method considers the response elicited at only a single frequency, and consequently, for most stimuli, remains limited in terms of information content. As an example in the context of visuotopic maps described above, current periodic models fail to provide any knowledge about the size of the underlying receptive fields’ population (pRF).

In this study, techniques developed for the analyses of fMRI-images and event related diffuse optical signal (Cohen-Adad et al., 2007, Marrelec et al., 2003) are used to model periodic optical signals by introducing a canonical hemodynamic response function (HRF). We will show that, by including higher harmonics in periodic techniques, this hemodynamic model provides a mean to extract additional information from the acquired data, such as HRF profile and pRF size. Results are validated in four steps: first, the hemodynamic model is validated with the “continuous stimulation paradigm” used to create visuotopic maps at a sub-millimeter scale of the cat primary visual cortex. Second, the linearity of the optical responses is investigated by combining multiple stimuli in one recording session. Third, multiple stimuli data is used to evaluate the profile of the hemodynamic response and compared to that obtained from episodic stimulation. Finally, harmonics are used to evaluate pRF size (i.e., second/first harmonic ratio). Part of this work has been previously published in abstract form (Vanni et al., 2007).

Section snippets

Signal model

The model presented in this work is based on the periodic stacking method (Sornborger et al., 2005) and is further developed by including a model of the HRF. Although the HRF can vary between animals and spatially in the cortex, it is useful to assume at first that it is known. A discussion of how to either estimate it or the limitations of the procedure will follow. The model will then be used to estimate responses and to show how higher harmonics can be used to infer new information from the

Visuotopic mapping of the visual cortex

As a first step to the analysis proposed in the previous section, we confirmed that visuotopic maps could be obtained in cats with the continuous paradigm. Fig. 2 displays the raw data used to compute these maps. The temporal profile of the raw signal recorded in one location of area 18 shows a major contribution of vasomotion (large periodic fluctuations observed in panel A). A slow drift is also observed that is probably due to instrumentation (light source and camera heating). By performing

Discussion

This study showed that the “continuous stimulation paradigm” developed by Kalatsky and Stryker (2003), can be successfully used to acquire visuotopic maps along the azimuth and elevation using vertical and horizontal bright bars in the cat primary visual cortex, with reduced acquisition time. More important perhaps is the evidence that, in the same recording session, up to four stimuli presented at different frequencies can be combined to evaluate the profile of the HRF. The latter was

Conclusion and future directions

Using continuous stimulation, clear visuotopic maps of the cat primary visual cortex were acquired in a short recording period and responses to mixed stimuli were simultaneously recorded in a given test. More important perhaps, the analysis of higher harmonics was exploited to estimate the receptive fields’ size of the underlying neurons. This measurement represents a new tool that will likely be valuable in projects requiring a thorough quantification of cortical organization (e.g.

Acknowledgments

This work was supported by NSERC and CIHR grants to C.C. and a NSERC grant to F.L. FRSQ provided part of C.C.’s salary (Chercheur National program). MV was supported in part by “Foreign Affairs and International Trade” and “Faculté des etudes supérieures” fellowships. JP was supported by a NSERC scholarship. We are grateful to J. Ribot for his comments on the manuscript.

References (47)

  • C. Olman et al.

    Retinotopic mapping in cat visual cortex using high-field functional magnetic resonance imaging

    J Neurosci Methods

    (2003)
  • A. Sornborger et al.

    Spatiotemporal analysis of optical imaging data

    NeuroImage

    (2003)
  • A. Sornborger et al.

    Extraction of the average and differential dynamical response in stimulus-locked experimental data

    J Neurosci Methods

    (2005)
  • M.P. Vanni et al.

    Bimodal modulation and continuous stimulation in optical imaging to map direction selectivity

    NeuroImage

    (2010)
  • M.Y. Villeneuve et al.

    Modular organization in area 21a of the cat revealed by optical imaging: comparison with the primary visual cortex

    Neuroscience

    (2009)
  • A. Zepeda et al.

    Optical imaging of intrinsic signals: recent developments in the methodology and its applications

    J Neurosci Methods

    (2004)
  • K. Albus

    A quantitative study of the projection area of the central and the paracentral visual field in area 17 of the cat. I. The precision of the topography

    Exp Brain Res

    (1975)
  • K. Albus et al.

    Second and third visual areas of the cat: interindividual variability in retinotopic arrangement and cortical location

    J Physiol

    (1980)
  • P.O. Bishop et al.

    Some quantitative aspects of the cat's eye: axis and plane of reference, visual field co-ordinates and optics

    J Physiol

    (1962)
  • G.G. Blasdel

    Differential imaging of ocular dominance and orientation selectivity in monkey striate cortex

    J Neurosci

    (1992)
  • T. Bonhoeffer et al.

    Optical imaging based on intrinsic signal – the methodology

  • W.H. Bosking et al.

    Spatial coding of position and orientation in primary visual cortex

    Nat Neurosci

    (2002)
  • W.H. Bosking et al.

    Functional specificity of callosal connections in tree shrew striate cortex

    J Neurosci

    (2000)
  • Cited by (0)

    View full text