|Title||Frequency transfer properties of the spike generating mechanism of cat retinal ganglion cells|
|Author(s)||Lankheet, M.J.M.; Molenaar, J.; De Grind, W.A. Van|
|Source||Vision Research 29 (1989)12. - ISSN 0042-6989 - p. 1649 - 1661.|
|Publication type||Refereed Article in a scientific journal|
|Keyword(s)||Cat - Frequency transfer properties - Ganglion cell - Retina - Spike generating mechanism|
The dynamic properties of the spike generating (SG) mechanism of retinal ganglion cells have been studied from intracellular recordings in the cat eye. Intracellularly recorded light flicker responses were separated by computer into spike trains and corresponding generator potentials. Both the spike train and the generator potential responses to temporally modulated light spots were analysed in terms of amplitude and phase plots. The differences in dynamic properties between the two response measures reveal that the SG-mechanism affects the temporal frequency transfer properties of retinal ganglion cells to a considerable extent. With respect to the transfer of the amplitude of the first harmonic the SG-mechanism has differentiating (or high-pass) properties. This means that the responses to high temporal stimulus frequencies are amplified relatively much more than are the responses to lower frequencies. Furthermore, the SG-mechanism causes a phase lead of the spike train response relative to the generator potential by, on average, 37 degrees. The measured frequency responses, among other things, have been used to verify and to quantify the SG-model that we proposed in a previous paper (Lankheet, Molenaar & van de Grind, 1989). With this model it proved possible to reproduce the spike train responses as model output from the corresponding measured generator potentials as model input. A good qualitative and quantitative correspondence between model output and the measured spike trains was obtained for a wide range of stimulus frequencies and with fixed values of the model parameters. With parameter values that optimized this correspondence the model allowed us to investigate the dynamic behaviour of the SG-mechanism in more detail. It also provides a reliable and validated method to predict the shape of the generator potential from the spike train (the "inversion problem").