Brain Advance Access originally published online on July 12, 2007
Brain 2007 130(11):e84; doi:10.1093/brain/awm154
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Reply: A few remarks on assessing magnocellular sensitivity in patients with schizophrenia
1Nathan Kline Institute for Psychiatric Research, Orangeburg, NY, 2Department of Psychiatry, New York University School of Medicine, New York, NY, 3City College of the City University of New York and 4Ferkauf Graduate School of Psychology, Yeshiva University, Bronx, NY, USA
Correspondence to: Pamela D. Butler, PhD, Nathan Kline Institute for Psychiatric Research, 140 Old Orangeburg Rd, Orangeburg NY 10962, USA E-mail: butler{at}nki.rfmh.org
Received April 24, 2007. Revised June 8, 2007. Accepted June 8, 2007.
Sir, Drs Skottun and Skoyles have written a critique of our paper ‘Subcortical visual dysfunction in schizophrenia drives secondary cortical impairments.’ We appreciate their interest in our paper and the opportunity to participate in scientific dialogue.
Our paper examines early visual processing in schizophrenia using two sets of stimuli. In Experiment 1 we used isolated checks of 4, 8, 16, 32 and 64% luminance contrast presented for 60 ms. The low contrast (4, 8%) stimuli were used to bias processing towards the magnocellular pathway, while the higher contrast stimuli were considered to activate both magnocellular and parvocellular pathways. In Experiment 2 we used high luminance contrast (80%) gabor patches (horizontal gratings with a gaussian envelope) of 1.0 cycle/degree (low spatial frequency: LSF) and 5.0 cycles/degree (high spatial frequency: HSF) presented for 100 ms. The LSF stimuli were used to bias processing towards the magnocellular pathway and the HSF stimuli were used to bias processing towards the parvocellular pathway.
The correspondents assert that because our stimuli in Experiment 2 were above threshold, they cannot be considered selective for the magnocellular or parvocellular pathway. They also criticize our use of low contrast in Experiment 1 to bias processing towards the magnocellular pathway and state that ‘under most conditions, it is the parvocellular, not the magnocellular system, that responds to the lowest contrast.’ Finally, they criticize our interpretation of our results. We gave the correspondents explicit feedback in response to a prior draft of their letter, including numerous literature citations, some of which they incorporated into their published critique and the rest they have chosen to ignore. We were and are aware of the literature they cite. Bridging the gap between perception or gross electrophysiology and the response characteristics of single neurons, much of which is known from studies on animals, necessarily involves speculation. Even given this latitude in educated opinions, however, we have to disagree with the arguments put forth by Skottun and Skoyles and feel that they have reached their conclusions in error.
First, we would like to point out that, contrary to the assertions of the correspondents, we never claimed that our stimuli completely isolated the magnocellular or the parvocellular system. No stimuli, including the ones advocated by the correspondents, can be expected to entirely segregate systems. We were therefore careful in our paper to use the terms ‘magnocellular-biased’ and ‘parvocellular-biased’ with regard to the stimuli, and have been careful to do so in our response. Also, as we noted in our original manuscript, it is critical to distinguish the subcortical magnocellular and parvocellular pathways from the cortical dorsal and ventral streams.
We would also like to point out that use of threshold-level stimuli, which are effective for psychophysical experiments, are impractical for event-related potential (ERP) studies. We (Butler et al., 2005
) and others (Slaghuis, 1998; Keri et al., 2002) have shown reduced contrast sensitivity, which by definition reflects contrast threshold, in schizophrenia. Such experiments, however, depend upon behavioural rather than electrophysiological measures. In the present study, the measure is amplitude of scalp-recorded electrical activity, which provides more specific information regarding neurophysiological response patterns in early visual regions. As shown by the 4% contrast data in Experiment 1, responses are extremely small to near threshold stimuli and signal-to-noise ratio is low. Thus, such stimuli are not optimal for ERP studies. Our stimuli in Experiment 1, which evaluated responses across a range of contrasts, allowed us to examine contrast gain control for the ERP response function. Contrary to the assertions of the correspondents, we feel that the non-linear nature of this function parallels the known magnocellular response function quite well. Neurons in other brain regions, such as MT, also receive magnocellular input and also show similar response functions. The high contrast LSF and HSF stimuli used in Experiment 2 provide robust responses which differentially activate dorsal versus ventral stream structures in normal volunteers, as predicted by their preferential activation of the magnocellular versus parvocellular pathways, respectively.
With regard to their arguments concerning spatial frequency, we feel that the correspondents mischaracterize the available data on two levels. First, we note that in humans, magnocellular cells have lower resolving ability than those found in monkeys, indicating that their responses are more biased toward lower spatial frequencies. This is because the dendritic fields of parasol (M) cells, but not midget (P) cells, are much larger in humans than in monkeys (Dacey and Petersen, 1992). In humans, at 3 degrees of eccentricity, M cells have a 10-fold greater dendritic field diameter than do P cells, which is expected to yield quite different receptive field sizes and spatial tuning functions. In central retina, the dendritic fields of M cells are about twice as large in humans as in macaques resulting in a 4-fold difference in area. Dacey and Petersen (1992) state that ‘this result predicts that the human parasol cells should show a lower resolving ability and an increased sensitivity to luminance contrast than their equivalents in the macaque.’ Thus, human M cells should exhibit lower spatial frequency tuning as well as higher contrast sensitivity than macaque M cells, and greater differences between M- and P-cell function would be seen in humans than in monkeys. In addition, other key differences have been observed in cortical recipients of magnocellular afferents in humans versus monkeys (e.g. Preuss and Coleman, 2002). Thus, absolute spatial frequency values from monkey studies cannot be applied blindly to the human literature.
Secondly, even within the monkey literature, the picture painted by Skottun and Skoyles is not fully accurate. They cite ‘characteristic frequencies’ from Levitt et al. (2001) of 3.55 and 4.57 cycles/degree, respectively, as indicating that the spatial frequency response patterns of magnocellular and parvocellular neurons are similar. This is a misleading quotation, as characteristic frequencies are very different from peak frequencies. Characteristic frequencies are the corner frequencies—the point at which the response of a given mechanism (e.g. receptive field centre) begins to fall off with regard to spatial frequency. Thus, this value is relatively uninformative with regard to response characteristics at the low spatial frequency end of the response range. The values cited by Levitt et al. (2001), however, do support our contention that responses to 5 cycles/degree stimuli (our HSF stimuli) would stimulate parvocellular neurons more than magnocellular neurons, which show a steep drop-off in monkeys over the 3.55 cycles/degree value quoted. In addition, in Levitt et al. (2001), the difference between the 3.55 and 4.57 cycles/degree characteristic frequencies is statistically significant (P < 0.02). The differences are likely even greater in humans based on the anatomical work of Dacey and Petersen (1992).
As opposed to characteristic frequency, there are other parameters in the monkey literature that do provide information about relative low spatial frequency response in magnocellular versus parvocellular neurons. For instance, in single-cell studies of monkey lateral geniculate nucleus (LGN), Derrington and Lennie (1984) report that magnocellular cells have peak responses at lower spatial frequencies than do parvocellular cells for a criterion well above threshold (Figures 8 and 13). In addition, Tootell et al. (1988b) using high contrast (70%) stimuli, found that HSF gratings produced much higher uptake of 2DG in 4Cß of primary visual cortex (which receives input from the parvocellular LGN layers), whereas LSF stimuli produced the opposite results with greater uptake in 4C
(which receives input from magnocellular LGN layers). Tootell et al. (1988b) state that ‘Presumably, cells in the magnocellular LGN layers and/or in the magnocellular-dominated layer 4C have lower average spatial frequency tuning (larger receptive fields) than their counterparts in the parvocellular LGN and/or in striate layer 4Cß.’
We also feel that the correspondents fail to cite the appropriate literature with regard to our use of isolated-check stimuli and contrast manipulation to bias activity toward the magnocellular pathway. In discussing contrast manipulation, the correspondents choose to ignore the same monkey single-cell studies that they cite with regard to spatial resolution (e.g. Kaplan and Shapley, 1982; Spear et al., 1994; Levitt et al., 2001). These studies show unequivocally that magnocellular neurons have greater contrast gain and respond at lower contrast than do parvocellular neurons. Cortical recipients of magnocellular input respond to low contrast whereas those receiving parvocellular input do not respond until contrast attains at least 
8% (Tootell et al., 1988a). Also, the correspondents do not cite the lesion literature accurately. Although they quote the literature as stating that ‘the largest deficits in contrast sensitivity in fact occur following parvocellular lesions,’ they neglect to mention that the finding is only for static stimuli (Merigan and Maunsell, 1993). The isolated-check stimuli that we used in the contrast experiment were presented with abrupt onset and offset for a brief duration of 60 ms, which corresponds to a transient condition with considerable high temporal frequency content. Also, the stimulus pattern had low spatial frequency composition (1.3 cycles/degree), as noted explicitly by the correspondents. Merigan et al. (1991a) explicitly state that ‘Magnocellular lesions greatly reduced detection contrast sensitivity at high temporal and low spatial frequencies,’ while the correspondents themselves point out that ‘reductions in contrast sensitivity following lesions in the magnocellular layers are mainly confined to cases where stimuli of both low spatial and high temporal frequency are employed (Schiller et al., 1990a, b; Merigan and Maunsell, 1990; Merigan et al., 1991a, b; Merigan and Maunsell, 1993).’ This is the situation that pertains to our stimuli.
The correspondents also cite Legge (1978) to say that the contrast threshold of a 3 cycles/degree grating ‘associated with the parvocellular system’ is 0.13%. However, the Legge (1978) statement pertains to stimuli of 3000 ms duration, which are much longer than the duration used in the present paradigm (
100 ms). As noted by the correspondents themselves, the length of stimulation is a critical issue in comparing data across studies, as integration time in the parvocellular (sustained) system is far longer than in the magnocellular (transient) system. In fact, Legge (1978) demonstrates that LSF conditions of 100 ms duration or less emphasize the transient system, which the correspondents themselves claim to be equivalent to the magnocellular pathway.
Thus, psychophysical work in both monkeys and humans, which includes lesion studies in monkeys, are in agreement that a low spatial frequency stimulus (e.g. 1 cycle/degree) presented at moderately high temporal frequency selectively activates the transient mechanism (Legge, 1978) and magnocellular pathway (Merigan et al., 1991a; Merigan and Maunsell, 1993) as we assert in our study.
We would also like to note that the isolated-check stimuli have been used in several previous VEP studies, most recently by Zemon and Gordon (2006). In these studies, the stimuli appear to have served their intended purpose, i.e. biased responses in favour of magnocellular or parvocellular activity depending on contrast manipulations. For instance, as contrast increases in the low to moderate range, the VEP response functions exhibit amplitude compression and phase advance for magnocellular-biased stimuli, which is consistent with responses obtained from M-type cells in the retina and lateral geniculate nucleus of monkeys (Derrington and Lennie, 1984; Kaplan and Shapley, 1986).
We appreciate that the correspondents have taken the time and effort to perform a Fourier analysis of our isolated-check stimuli. Contrary to their claims, however, their analysis actually supports our contention that these stimuli are magnocellularly biased. We claimed in our article that isolated-check stimuli were magnocellularly biased at low contrast. As they show, the main Fourier components are at a low spatial frequency (1.3 cycles/degree) and the secondary higher spatial frequency components have greatly reduced energy, and thus would not be expected to contribute much to the response at low contrast. It is quite possible that these stimuli are biased toward the magnocellular system even at higher contrasts.
The correspondents miscite our article in criticizing our conclusions. Their critique is based upon the assertion that ‘Butler et al. (2007) associate the ventral cortical stream with parvocellular activity ...’ This is untrue. In fact, in our paper (p. 418) we state quite clearly that there is ‘some convergence of magnocellular and parvocellular input even in V1 (Sawatari and Callaway, 1996; Vidyasagar et al., 2002) and significant interaction between dorsal and ventral streams occurring thereafter.’ Thus, in particular, the ventral stream receives direct input from the parvocellular pathway, and crossover input from the magnocellular pathway via secondary visual regions (e.g. V3A) (Chen et al., 2006). Only the correspondents incorrectly equate the subcortical parvocellular pathway and the cortical ventral stream pathways.
In their critique they ignore the major finding of our study, i.e. that the decreased C1, P1 and N1 amplitudes in the transient VEPs are found only to LSF, not HSF, stimuli, as shown in Table 3 of the paper. In our study, many patterns of results were possible. We could have observed normal activity, in which case we would have concluded that early visual processing in schizophrenia was intact. We could have observed reduced C1, P1, or N1 activity regardless of stimulus type, in which case we would have concluded that the cortical region or mechanism generating that component was specifically affected. What we observed, however, is that responses starting with C1 and continuing with P1 and N1 were deficient to LSF, but not HSF, stimuli. We concluded that the most parsimonious explanation for this finding is that the subcortical pathway providing LSF input to cortex (i.e. magnocellular pathway), rather than cortex itself, was impaired. This conclusion was buttressed by the fact that patients also showed reduced ERP generation to low (4, 8%) contrast stimuli. Such stimuli also bias activity toward the magnocellular pathway.
Finally, some assertions by the correspondents simply confuse us. For example, they state that ERP deficits, because they measure cortical activity, ‘could be the result of deficiencies anywhere along the visual pathways (i.e. in the visual optics, photoreceptors, retinal ganglion cells, LGN cells) ...’ This is certainly true, although the last two nodes (retinal ganglion cells, LGN cells) are components of the magnocellular pathway and so would fall within our hypothesis. It is hard to see how difference in optics would yield the results that we obtained. With regard to optics, we note that all subjects were corrected to vision of 20/30 or better. Under our photopic conditions, the same photoreceptors (cones) contribute to the M- and P-pathways and thus are unlikely to cause a differential deficit.
These issues were all pointed out to the correspondents in our response to the previous draft of their letter. We note that they did make some corrections to their original letter in response to our initial comments (e.g. Skottun and Skoyles had originally stated that there was no low spatial frequency attenuation in parvocellular response functions). However, they have failed to either acknowledge or refute most of our comments, but continue to make their assertions. To us, it seems that they would be better served by conducting original studies using stimuli that they feel are optimal.
We are gratified by the correspondents’ statement that visual pathway function is an important area of study in schizophrenia and that magnocellular dysfunction would have ‘significant implications’ for understanding schizophrenia. We also believe that scientific exchange is extremely important. We wish, however, that Drs Skottun and Skoyles had examined a broader literature prior to formulating their critique, and that they had cited the extant literature, as well as our article, more accurately. We hope that the literature reviewed and the arguments put forth here will be useful to others in designing experiments. As always, our hypotheses, like all scientific hypotheses, are testable. Further work is needed to replicate our findings as well as to provide a greater understanding of visual dysfunction and its implications in schizophrenia and other disorders.
 |
References |
|---|
 Top References |
|---|
Butler PD, Martinez A, Foxe JJ, Kim D, Zemon V, Silipo G, et al. Subcortical visual dysfunction in schizophrenia drives secondary cortical impairments. Brain (2007) 130:417–30.
Butler PD, Zemon V, Schechter I, Saperstein AM, Hoptman MJ, Lim KO, et al. Early-stage visual processing and cortical amplification deficits in schizophrenia. Arch Gen Psychiatry (2005) 62:495–504.
Chen CM, Lakatos P, Shah AS, Mehta AD, Givre SJ, Javitt DC, et al. Functional anatomy and interaction of fast and slow visual pathways in macaque monkeys. Cereb Cortex (2006).
Dacey DM, Petersen MR. Dendritic field size and morphology of midget and parasol ganglion cells of the human retina. Proc Natl Acad Sci USA (1992) 89:9666–70.
Derrington AM, Lennie P. Spatial and temporal contrast sensitivities of neurones in lateral geniculate nucleus of macaque. J Physiol (1984) 357:219–40.
Kaplan E, Shapley RM. X and Y cells in the lateral geniculate nucleus of macaque monkeys. J Physiol (Lond) (1982) 330:125–43.
Kaplan E, Shapley RM. The primate retina contains two types of ganglion cells, with high and low contrast sensitivity. Proc Natl Acad Sci USA (1986) 83:2755–7.
Keri S, Antal A, Szekeres G, Benedek G, Janka Z. Spatiotemporal visual processing in schizophrenia. J Neuropsychiatry Clin Neurosci (2002) 14:190–6.
Legge G. Sustained and transient mechanisms in human vision: temporal and spatial properties. Vision Res (1978) 18:69–82.[CrossRef][Web of Science][Medline]
Levitt JB, Schumer RA, Sherman SM, Spear PD, Movshon JA. Visual response properties of neurons in the LGN of normally reared and visually deprived macaque monkeys. J Neurophysiol (2001) 85:2111–29.
Merigan WH, Byrne CE, Maunsell JH. Does primate motion perception depend on the magnocellular pathway? J Neurosci (1991a) 11:3422–9.[Abstract]
Merigan WH, Katz LM, Maunsell JH. The effects of parvocellular lateral geniculate lesions on the acuity and contrast sensitivity of macaque monkeys. J Neurosci (1991b) 11:994–1001.[Abstract]
Merigan WH, Maunsell JH. Macaque vision after magnocellular lateral geniculate lesions. Vis Neurosci (1990) 5:347–52.[Web of Science][Medline]
Merigan WH, Maunsell JHR. How parallel are the primate visual pathways? In: Ann Rev Neuroscience.—Cowan WM, Shooter EM, Stevens CF, Thompson RF, eds. (1993) Palo Alto, CA: Annual Reviews, Inc. 369–402.
Preuss TM, Coleman GQ. Human-specific organization of primary visual cortex: alternating compartments of dense Cat-301 and calbindin immunoreactivity in layer 4A. Cereb Cortex (2002) 12:671–91.
Sawatari A, Callaway EM. Convergence of magno- and parvocellular pathways in layer 4B of macaque primary visual cortex. Nature (1996) 380:442–46.[CrossRef][Medline]
Schiller PH, Logothetis NK, Charles ER. Functions of the colour-opponent and broad-band channels of the visual system. Nature (1990a) 343:68–70.[CrossRef][Medline]
Schiller PH, Logothetis NK, Charles ER. Role of the color-opponent and broad-band channels in vision. Vis Neurosci (1990b) 5:321–46.[Web of Science][Medline]
Slaghuis WL. Contrast sensitivity for stationary and drifting spatial frequency gratings in positive- and negative-symptom schizophrenia. J Abnorm Psychol (1998) 107:49–62.[CrossRef][Web of Science][Medline]
Spear PD, Moore RJ, Kim CB, Xue JT, Tumosa N. Effects of aging on the primate visual system: spatial and temporal processing by lateral geniculate neurons in young adult and old rhesus monkeys. J Neurophysiol (1994) 72:402–20.
Tootell RB, Hamilton SL, Switkes E. Functional anatomy of macaque striate cortex. IV. Contrast and magno-parvo streams. J Neurosci (1988a) 8:1594–609.[Abstract]
Tootell RB, Silverman MS, Hamilton SL, Switkes E, De Valois RL. Functional anatomy of macaque striate cortex. V. Spatial frequency. J Neurosci (1988b) 8:1610–24.[Abstract]
Vidyasagar TR, Kulikowski JJ, Lipnicki DM, Dreher B. Convergence of parvocellular and magnocellular information channels in the primary visual cortex of the macaque. Eur J Neurosci (2002) 16:945–56.[CrossRef][Web of Science][Medline]
Zemon V, Gordon J. Luminance-contrast mechanisms in humans: visual evoked potentials and a nonlinear model. Vision Res (2006) 46:4163–80.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||