Temporal Frequency and the Magnocellular and Parvocellular Systems

Abstract

Can the magnocellular system be stimulated exclusively, or predominantly, by using a particular temporal frequency? There are two problems. (1) Many researchers have confused contrast reversal rates with luminance modulation frequencies, even though it is the latter (which is half the reversal rate) that may be compared to neuronal responses. (2) Many of the frequencies used are too low to selectively activate magnocellular neurons. While a pure magnocellular response may be obtained by employing particular combinations of temporal frequency, spatial frequency and contrast, however, this possibility is limited to stimuli close to detection threshold, and such stimuli give only weak responses.

KEYWORDS:

• Magnocellular system
• contrast reversal rate
• luminance modulation frequency
• selective activation

Notes

*Johansson and Jakobsson⁶ wrote that “[t]he visual system is stimulated and a response … is evoked each time the pattern is reversed.” While this may be strictly correct as far as overall VEPs are concerned it does not mean that the individual neurons were in fact stimulated at this rate. The above statement of Johansson and Jakobsson⁶ would be misleading if it were taken to mean that the response properties of individual cells given in terms of temporal frequency ought to be compared to the reversal rate of the stimulus. The temporal characteristics of single magno- and parvocellular neurons need to be compared in terms of modulation frequency, not the reversal rate.

**Spear et al.¹³ measured temporal cutoff frequency by fitting regression lines to the data on the high frequency slope of the temporal frequency response functions and then noting at which frequency this line intersects 2 SE above the inherent F1 activity measures during the presentation of a blank screen. Levitt et al.¹⁴ measured cutoff by determining the temporal frequency (at the high frequency slope of the function) at half peak response. In the study of Movshon et al.¹⁵ the temporal frequency response data were fitted with curves of the form R(w_t) = (k exp[(− w_t/f_c)²]/[1+ (f_h/w_t)^b] where w_t is the temporal frequency, k is a scaling constant, f_c is the characteristic frequency, f_h sets the corner frequency of the low-frequency limb, and b sets the slope of the low frequency limb. The characteristic frequency, f_c, was taken as the measure of temporal resolution.

***Slaghuis and Ryan²⁸ made this claim on the basis of Burbeck and Kelly.³² In that study psychophysical observations were related to neurophysiology in terms of X- and Y-cells, not parvo- and magnocellular neurons. The X/Y dichotomy was first made in the cat,³³ a species which does not have magno- and parvocellular systems. The X- and Y-cell classification in cats cannot be related to the magno-/parvocellular distinction in a direct way. (While essentially all parvocellular neurons are X-cells some–about 5% −33%–of the magnocellular neurons have response properties consistent with Y cells.¹⁴)