Global Shape Processing Deficits Are Amplified by Temporal Masking in Migraine. Doreen Wagner, Velitchko Manahilov, Gael E. Gordon, and Gunter Loffler

Transcription

1 Visual Psychophysics and Physiological Optics Global Shape Processing Deficits Are Aplified by Teporal Masking in Migraine Doreen Wagner, Velitchko Manahilov, Gael E. Gordon, and Gunter Loffler PURPOSE. Individuals with igraine show subtle defects in a range of visual tasks copared to nonigraineurs. Increased neuronal noise can account for soe of these deficits. To exaine the generality of increased noise in igraine, asking effects were copared in igraineurs and headache-free controls using a shape discriination task, thought to involve processing in extrastriate cortical areas. METHODS. Nine igraineurs with aura, nine igraineurs without aura, and nine headache-free controls participated. Observers had to detect deviations in circular shapes with or without a larger contour ask. The nonoverlapping ask was presented at five teporal intervals (stiulus onset asynchronies, SOA): 0 (siultaneous), 66, 100, 133, and 250 s. RESULTS. Migraineurs with aura perfored worse in all tests than igraineurs without aura and controls. Both igraine groups perfored poorer than controls at discriinating shapes without asks. Typical asking functions were obtained fro all groups, but they were steeper for igraineurs than controls with thresholds raised ost draatically (2.1 and 4.4 ties for igraineurs without and with aura relative to controls, respectively) at SOAs where asks had their ost detriental effect ( s). Modeling the effect of asking showed that raised internal noise alone is insufficient to explain these deficits. Rather, an abnoral nonlinear transducer function (e.g., as part of gain-control) together with increased ultiplicative noise is required to capture the data. CONCLUSIONS. The findings are consistent with an extrastriate deficit in igraine that cannot be explained copletely by defective inhibition. (Invest Ophthalol Vis Sci. 2013; 54: ) DOI: /iovs Migraineurs, especially those with aura, are poorer than controls at discriinating shapes. This deficit is aplified drastically when shapes are presented in the context of a teporal ask. 1 Migraine auras are thought to result fro cortical spreading depression (CSD), a wave of suppressed neuronal activity in the visual cortex. 2 It has been hypothesized that interictal cortical hyperexcitability ight leave igraine patients vulnerable to the spontaneous neuronal depolarization that characterizes CSD. 3 One possible echanis for cortical hyperexcitability is a lack of inhibitory control, 4 resulting perhaps fro ipaired Fro the Departent of Life Sciences, Glasgow Caledonian University, Glasgow, United Kingdo. Subitted for publication Noveber 2, 2012; revised Deceber 26, 2012; accepted Deceber 29, Disclosure: D. Wagner, None;V. Manahilov, None; G.E. Gordon, None; G. Loffler, None Corresponding author: Gunter Loffler, Departent of Life Sciences, Glasgow Caledonian University, Glasgow, UK; g.loffler@gcu.ac.uk. gaa-ainobutyric acid (GABA)-ergic inhibitory echaniss 5 or a deficiency of inhibitory neurons. 1 This reducedinhibition hypothesis has been tested in a nuber of studies. Paler et al. easured the detriental effect of a nonoverlapping ask presented shortly after a target 6 (etacontrast asking 7 ). It has been suggested that the asking behavior results fro a fast transient signal, triggered by the ask, inhibiting a slower sustained target-related activation. 8 Reduced cortical inhibition, therefore, would be expected to result in reduced asking. Consistent with this, Paler et al. found that igraine with aura subjects perfored better during etacontrast asking than either headache-free controls or igraine without aura subjects. 6 Recently, Shepherd et al. confired that igraineurs experienced less asking than controls. 9 However, igraineurs also discriinated between targets better when they were unasked, iplying that a heightened neuronal response in igraine ay underlie their enhanced perforance, rather than a specific deficit with inhibitory processes. Other easures of inhibitory control have found igraineurs to perfor at a near-noral 10,11 or reduced level, 12,13 also inconsistent with the faulty inhibition odel. An alternative explanation for perceptual differences between igraineurs and controls can be based on increased internal noise, a possible consequence of cortical hyperexcitability. 14 Migraineurs perfor worse than controls in any tasks, consistent with increased levels of internal noise. Differences are reported for low-level tasks, including discriination of spatial frequency, 15 orientation, 11 flicker, 14 color, 16,17 luinance, 18,19 and long-range inhibition, 20 as well as tasks requiring extrastriate processing, such as global shape 21,22 and global otion coherence thresholds. 14,21 23 Coparison of internal noise levels has shown differences between igraine and controls. 18,19,24 The ai of our study was to investigate the nature and extent of visual deficits in igraineurs, and to distinguish between two possible factors: abnoral inhibition and abnoral internal noise. The task required the discriination of circular shapes in the absence and presence of a ask. Huan sensitivity for detecting subtle deforations of circular shapes is in the hyperacuity range. 25 Psychophysical, onkey physiology, 29 and huan fmri 30 studies support the view that the high sensitivity ay be a result of processing at interediate, extrastriate stages. 31 According to the reduced inhibition hypothesis, if asking was a consequence of inhibition, igraineur perforance should be siilar to controls without a ask, but better than controls when asked. If, however, igraineurs have raised internal noise levels, they should perfor equally poorly in both conditions. Our results supported neither prediction. Migraineurs (especially those who experience visual auras) perfored slightly worse than controls in the absence of a ask, but this difference was aplified significantly by the presence of a ask. Investigative Ophthalology & Visual Science, February 2013, Vol. 54, No Copyright 2013 The Association for Research in Vision and Ophthalology, Inc.

2 IOVS, February 2013, Vol. 54, No. 2 Global Shape Processing Deficits in Migraine 1161 METHODS Participants Headache-free controls were recruited at Glasgow Caledonian University, Glasgow, UK. The control group had never experienced a igraine and reported no ore than one headache a onth. This was verified by a structured questionnaire in accordance with the second edition of the International Headache Classification (ICHD-II). 32 Migraineurs were recruited at the Middle-Geran Headache Centre, Neurology Departent, University Clinic Jena, Jena, Gerany. For subjects to be included as igraineurs, they were required to fulfill the classification criteria of the International Headache Society. Diagnosis was undertaken by a neurologist. Migraineurs were divided into two groups: igraine with aura (MA) or igraine without aura (MO). Subjects in the igraine groups also copleted the Migraine Disability Assessent (MIDAS) questionnaire. 33 A total of 27 subjects participated in this study: 9 headache-free control subjects (ean age 6 SD, years), 9 MA ( ), and 9 MO ( ). The sex distribution in each group was 7:2 (feale-to-ale ratio). The igraine groups took no preventative edication in the three onths before the study, and had no igraine episodes during the 3 days before or after the experient. Median headache frequency in the onth before the study was 5.0 and 2.0 for MO and MA, respectively, and did not differ between the two groups (2 tailed t-test, P ¼ 0.330). Median MIDAS scores were 16.0 and 11.0 for MO and MA, and were not significantly different (2 tailed t-test, P ¼ 0.235). Migraine participants were ranked for the duration of their igraine (1 2, 2 5, and >5 years); the two groups did not differ fro each other (Wilcoxon signed rank test, P > 0.05). All subjects had noral or corrected-to-noral vision, and wore their habitual correction. Each participant et the following ocular criteria: visual acuity of 20/30 or better, intraocular pressures of less than 21 Hg (Goldann applanation tonoetry), and noral visual fields (Huphrey 30-2 or Oculus fast threshold). Exclusion criteria were pregnancy, age above 40 years, epilepsy, dyslexia, and diabetes. All subjects gave written infored consent in accordance with a protocol approved by the ethics coittees of Glasgow Caledonian University and the University Clinic Jena, in agreeent with the tenets of the Declaration of Helsinki. Apparatus Stiuli were presented on a high-resolution coputer screen (Iiyaa Vision Master Pro 450; Iiyaa International, Oude Meer, The Netherlands) with spatial resolution of pixels and 120 Hz refresh rate, fro a distance of 85 c. Viewing was binocular. The roo was seidarkened ( cd/ 2 ) and the ean luinance of the screen was cd/ 2. A custo-ade video suation device 34 was used to give 12-bit precision. The onitor gaa nonlinearity was calibrated and verified regularly with an OptiCAL photoeter (Cabridge Research Systes, Ltd., Rochester, Kent, UK). Stiuli The study used a set of radial frequency (RF) patterns, 25 which are defined as a sinusoidal odulation to the radius of a circle in polar coordinates (Fig. 1). Different shapes can be generated by varying the frequency (nuber of lobes) or aplitude (sharpness of each lobe) of the sinusoid. The shape used throughout this study was an RF5 (i.e., fivesided, pentagon-like shape). The aplitude of the RF5 was anipulated to obtain shape discriination thresholds, that is the iniu aplitude required to discriinate reliably between an RF5 and a circle. Procedure Participants received detailed instructions before an experiental session. A teporal two-alternative forced choice paradig was used: observers were required to report which of two target shapes, presented sequentially, appeared less circular. The screen was set initially to a unifor grey field of ean luinance. A fixation point appeared at the center of the screen, followed by target shapes presented briefly (25 s, 3 fraes) either in isolation (without asking), with a siultaneous ask, or followed by a ask of equally short duration (backward asking). Shape discriination was easured for five different stiulus-onset asynchronies (SOAs; 0 ¼ siultaneous asking, 66.7, 100, 133.3, and 250 s). Different SOAs were run in separate blocks. The order in which observers ran blocks was randoized. During the SOA period, the screen returned to the unifor grey field. Before the testing coenced, subjects were given the opportunity to failiarize theselves with the setup. No feedback was provided during practice period or data collection. Each observer ran all conditions. Statistical Analysis Before group analyses, individual data were copared to group eans to identify outliers that are three or ore standard deviations fro the ean. None of the data fell outside this range. Data for the no-ask condition were copared using an ANOVA with subject group (igraine, MA, MO) as factor. Data for the asked condition were copared using a repeated-easures ANOVA with ask tiing (SOAs of 0, 66, 100, 133, and 250 s) and group (igraine, MA, MO) as factors. All subsequent pairwise coparisons were done using t-tests with Bonferroni adjustents for ultiple coparisons. RESULTS Shape Discriination Thresholds for shape discriination were easured as the iniu aplitude of the RF pattern required to discriinate it fro a circle and they are expressed in percent of the radius of the circle. Without asking, the average sensitivity (6SD) for discriinating between a circle and an RF5 (pentagon-like shape) for the control subjects was % (Fig. 2). Hence, noral observers can just discriinate between the two left-ost shapes in Figure 1A. Migraineurs perfored poorer than controls: average thresholds were % and for MO and MA, respectively. There was a significant ain effect of subject condition (F (2,24) ¼ 3.58, P < 0.05). MA perfored significantly (P < 0.05) poorer than controls. MA thresholds were 2.2 ties higher than the control group. A typical MA can just discriinate between the contour on the left hand side of Figure 1A and the third shape fro the left. Average thresholds for MO were raised 1.37 ties copared to controls, but this difference did not reach significance. The difference between MO and MA also was not significant. Shape Discriination in the Presence of a Mask The second experient investigated the effect of asking on shape discriination. Figure 3 presents the data for the three groups at the five different SOAs (0, 66.7, 100, 133.3, and 250 s). Backward asking studies have shown that the detriental effect of a ask on a target depends on the SOA, producing a typical inverted U-shaped function. 35 Little asking is seen when target and ask are presented siultaneously (SOA ¼ 0) or when they are sufficiently separated in tie. The data (Fig. 3) show the typical asking pattern 36 for all three groups with significantly raised thresholds for SOAs between 66 and 100 s. For controls, a peak asking effect between 66 and 100 s is in broad agreeent with Habak et al., who reported a peak between 80 and 110 s for the three norals tested in their study. 36 Repeated-easures ANOVAs

3 1162 Wagner et al. IOVS, February 2013, Vol. 54, No. 2 FIGURE 1. Shape stiuli and asking paradig. Shape discriination thresholds were easured for continuous RF patterns with five lobes. Observers had to discriinate between test shapes of varying aplitudes and a reference (circle). The cross-sectional luinance profile of the shapes was defined by a fourth spatial derivative of a Gaussian and set to a peak spatial frequency of 8 cpd. Contrast was 90%. (A) Shapes are shown with different aplitudes ranging fro zero (circle) to 40% odulation relative to the ean radius. Typical observers require an aplitude of 0.5% to discriinate a circle fro an RF5 contour. Migraine with aura (MA) were less sensitive for this task and showed thresholds of approxiately 1.5%. When the stiuli to be discriinated are followed by asks (see [B]), all observers becae less sensitive. For the ost disruptive ask tie (66 s), thresholds for controls were at approxiately 15%. For MA, they were at approxiately 40%. Hence, MA can only just distinguish between the shapes at the far right and far left in (A) when asked. (B) Experiental paradig. Observers were presented with two targets and had to indicate which of the two contained the noncircular shape (Interval II in this exaple). The aplitude of the noncircular target was under the control of the coputer in a staircase procedure to deterine thresholds. Shapes were presented briefly (25 s) either in isolation or in the presence of a ask. The ean radius of the shapes to be discriinated (targets) was 18. When present, the ask always was bigger in size than the target (1.58) to avoid spatial overlap between target and ask. The aplitude of the ask was set at 16 ties the detection threshold of an RF5 against a circle and, therefore, was visible clearly as a noncircular shape. Pattern orientation was assigned randoly fro trial to trial, but the orientations of the ask and test always were locked in-phase. This and the ask aplitude yield axiu asking effects for RF shapes. Thresholds were easured for different onset ties of the ask relative to the target (SOA). Masks were shown for the sae short duration as the targets. found significant ain effects of SOA for controls (F (4,32) ¼ 7.07, P < 0.001), MO (F (4,32) ¼ 4.20, P < 0.01), and MA (F (4,32) ¼ 10.23, P < 0.001). For all three groups, pairwise coparisons showed significantly (P < 0.05) elevated thresholds for SOA ¼ 66 s copared to SOA ¼ 0 s and SOA ¼ 250 s. For MA, differences between SOA ¼ 66 s and SOA ¼ 133 s also were significant. Thresholds for noral observers were elevated 6.3 ties for an SOA of 66 s relative to the siultaneous condition (SOA ¼ 0 s). The two igraine groups showed the sae general pattern of asking and the sae tiing for the peak asking effect. However, the agnitude of the peak asking effect was substantially greater in igraine. The peak threshold elevation was 11.8 ties for MO and 21.6 for MA. To assess differences between groups, data were analyzed by a two-way (3 3 5) repeated easures ANOVA with subject group (controls, MA, MO) and asking SOA (0, 67, 100, 133, and 250 s) as ain factors. This revealed significant ain effects for both factors: subject group (F (2,24) ¼ 3.72, P < 0.05) and SOA (F (4,96) ¼ 17.92, P < 0.001). There also was a significant interaction between the two factors (F (8,96) ¼ 3.12, P < 0.05). The differences between controls and MA, but not between MO and MA or MO and controls, were significant. At the peak SOA (66 s), the MA group perfored significantly poorer than the other two groups. Thresholds for the MA group were higher by factors of 2.1 and 4.4 copared to MO and controls, respectively. Those for MO were raised 2.1-fold copared to controls. The finding that igraineurs perfored poorer than controls under asking conditions is contrary to what has been reported for eta-contrast asking. As has been pointed out recently, 9 care should be taken when considering differences between groups for backward asking conditions when these groups also behave differently in the absence of a ask. Accordingly, we noralized the data for the asked conditions (Fig. 3) by the baseline sensitivity for the no-ask condition (Fig. 2) for each observer. The resulting threshold elevations are plotted in Figure 4. Differences between groups still were evident for the ost disruptive SOA (66 s). This difference reained significant for MA copared to controls (P < 0.05). For the other SOAs, the ain effect of clinical group did not reach significance (F (2,24) ¼ 1.33, P > 0.05). This suggests that asking per se disrupts processing in MA and

4 IOVS, February 2013, Vol. 54, No. 2 Global Shape Processing Deficits in Migraine 1163 FIGURE 2. Shape discriination thresholds for discriinating between a circle and an RF5. Thresholds are defined as the aplitude of the RF5 pattern at which observers could discriinate it reliably fro a circle and given relative to the size of the pattern (in percent). Data for the control group (left hand bar) averaged %, which is in the hyperacuity range. MO (right hand bar, average thresholds of %) perfored worse than the control group, but this does not reach significance. Indicated by the asterisk, however, thresholds for MA were significantly elevated ( %) copared to controls (P < 0.05). Black circles show thresholds for individual participants. Error bars here and elsewhere are standard errors of the ean. decreases further the already reduced overall shape discriination sensitivity in igraine without ask. Model Predictions We copared various odel predictions to the data in an attept to identify the source of the deficits in igraine. The FIGURE 4. Noralized data for shape discriination in the presence of a ask. Data for each observer fro Figure 3 were noralized against their respective thresholds for the condition without a ask. The difference between groups decreased because the two igraine groups showed reduced general sensitivity for the baseline condition (without ask, Fig. 2), but did not disappear. For the SOA where peak asking occurred, threshold elevations are significantly higher for MA copared to controls. *P < odel used here is a odified version of the Perceptual Teplate Model (PTM) 37,38 that has been used widely to quantify the aount of various internal noise sources. The odel (Fig. 5, Appendix) uses a divisive, inhibitory gain control process, which reduces the response of the target. 39 Gain control echaniss noralize the response of a neuron with regards to, for exaple, contrast, 40,41 consistent with physiologic and psychophysical evidence. 42 The consequences of gain control have been linked to the effect of asking on target visibility. 39,40,43 Figure 5A illustrates the odel in the case of a FIGURE 3. Shape discriination in the presence of a ask. Thresholds present the average for each of the three groups as a function of target-ask SOA. All groups exhibited the typical asking pattern with substantial threshold elevations for backward asking, and peak asking effect for SOAs between 66 and 100 s. Copared to controls, both igraine groups experienced a significantly stronger general asking effect with poorer overall perforance. Differences between groups were ost proinent where the ask for all groups exerts its strongest effect (SOA of 66 s). The asterisk indicates that the thresholds for MA were elevated significantly copared to each of the other two groups.

5 1164 Wagner et al. IOVS, February 2013, Vol. 54, No. 2 FIGURE 5. The structure of the gain control odel (A) and odel predictions for various paraeter anipulations (B E). (A) Target and ask stiuli are processed by perceptual teplates with gains of b and b, and their outputs are subject to nonlinearities (c and c ). Two internal noise sources occur before and after the gain control stage. The response to the target stiulus is inhibited by a divisive gain control process, which consists of the cobined activity to the target and ask. Finally, stiulus discriinability is calculated by the total signal-to-noise ratio (see Appendix). (B E) Model predictions. The dotted and dashed curves show the data for the control group and the MA group, respectively (for clarity the MO data are not shown). To show the effect of different odel paraeters on the asking profiles, paraeters initially were set to capture the control group s behavior (b ¼ 5, c ¼ 1.8, b ¼ f(soa) ¼ [3, 20, 18, 9, 4.8] for SOA ¼ [0, 66, 100, 133, 250 s], c ¼ c ¼ 1.8, N 1 ¼ , N 2 ¼ 0.1). (B E) The solid curves show how the control s thresholds would be affected when one individual odel paraeter is varied ([B]: for internal noise N 1 ¼ 0.06; [C]: N 2 ¼ 0.5; [D]: exponent of the nonlinear transducer function for the target c ¼ 3; [E]: exponent of the nonlinear transducer function for the ask c ¼ 3.3). See text for further details. target and a ask stiulus. The gain control in this odel is related to the interactions between shape encoding processes, that is neurons selective for the target and the ask. Evidence supports this type of interaction at higher stages of the ventral pathway. 29,31,36,44 Poorer perforance in igraine can be explained in a nuber of ways. An increase in either of the two internal noise sources (N 1 or N 2, pre- and post-gain control, respectively) will cause thresholds to rise. Siilarly, a less efficient perceptual teplate (b) or a different nonlinear transducer function (e.g.,

6 IOVS, February 2013, Vol. 54, No. 2 Global Shape Processing Deficits in Migraine 1165 an increased exponent of the power function) also would predict lower sensitivity. The way in which the asking curves (Figs. 3, 4) for the three groups differ provides an opportunity to distinguish between these possibilities. Initially, the gain of the ask (b ), which is a function of tie (SOA), was set to fit the data for the control group (dotted-line labeled controlodel in Fig. 6B). b is the odel paraeter that produces the inverted U-shaped asking profile. The effect of various odel paraeters on the shape of the asking function then can be investigated by coparing this original asking curve (Fig. 6B, dotted curve) to those resulting fro individual paraeter changes (Figs. 5B E, solid curves). Increasing the variance of the internal noise source that precedes the divisive inhibition (N 1 ) raises thresholds for short and long SOAs without affecting the peak SOA (Fig. 5B). This causes the asking function to flatten, which fails to capture the perforance of the MA group (dashed line). Increasing the variance of the internal noise that follows the divisive inhibition (N 2 ) predicts a general upward shift of thresholds (Fig. 5C). This shift is unifor across SOA and results in constant threshold elevations, also inconsistent with the MA data. Reducing the gain (b) results in the sae prediction. Increasing the exponent of the nonlinear transducer function for the target (c) has a siilar effect to increasing the N 1 noise (Fig. 5D). It predicts threshold elevations that disproportionally affect long and short SOAs where the ask has a relatively inor effect, resulting in an overall flattening of the asking curve. Finally, increasing the exponent of the nonlinear transducer function in the gain control pathway (c ) disproportionally affects the peak SOA, causing the asking curve to becoe steeper (Fig. 5E). This anipulation results in a curve that closely follows the overall shape of the MA group, but underestiates absolute thresholds. It follows that the only odel paraeter that predicts that thresholds for peak SOAs are ost draatically elevated in igraine is the nonlinearity in the gain control network (c 2 ). This can be cobined with an increase in internal noise (N 2 ), which by itself causes threshold to be elevated uniforly, to capture the specific deficits in igraine. To provide quantitative odel predictions, we fixed all other paraeters on the data for the controls. The data for MA and MO without ask then were used to deterine N 2 (Fig. 6A) and the backward asking curves to deterine c (Fig. 6B). The odel provides a satisfactory prediction of the data. DISCUSSION Migraineurs, especially those with aura (MA), show lower sensitivity than controls for closed contour shape discriination. This difference was not found in a recent study with sapled shapes. 19 In that study, external noise was added to the saples and internal noise levels estiated. Consistent with our analysis, increase in internal noise was present in igraine (although the additive internal noise was raised). However, this was coupled with an increased efficiency offsetting increased noise levels and resulting in the sae overall perforance. Reduced sensitivity in igraine observed in our study, however, is in line with earlier reports on global for perception, 22 where igraineurs were poorer at detecting a circular texture than headache-free controls. Like our shape tests, these tasks require global pooling and are consistent with a deficit in extrastriate ventral strea in igraine. It ay be argued that early processing deficits (e.g., processing of local contour orientation in V1) could account for lower sensitivity for tasks targeting higher stage processing. This possibility has been ruled out for global for processing in FIGURE 6. Model predictions. (A) The data for the shape discriination task in the absence of a ask were fitted with the odel (see Appendix). Model paraeters were set to fit the data for the control group (b ¼ 5, b ¼ f(soa) ¼ [3, 20, 18, 9, 4.8] for SOA ¼ [0, 66, 100, 133, 250 s], c ¼ c ¼ 1.8, N 1 ¼ , N 2 ¼ 0.1). All but N 2 then were fixed, and values for N 2 deterined that fit the data for MO (N 2 ¼ 0.3) and MA (N 2 ¼ 0.63). (B) Backward asking data for the three groups, including odel predictions (controls: c ¼ 1.8; MO: c ¼ 2.8; MA: c ¼ 3.3). See text for details. igraine 45 on the basis of noral processing at the level of V1. Early deficits also are unlikely to explain igraineurs perforance in this study. The asking effect for RF shapes is highly contextual. Masking strength depends on the relationship between the orientations of the ask and target, 36 and such a global-shape specific effect is inconsistent with V1 processing. Moreover, asking of V1 processes (eta-contrast asking) results in igraineurs perforing better than controls, 6,9 opposite to the effect described here. When target shapes are backward asked, igraineurs, especially MA, perfored poorer than controls. Although all groups exhibit the typical inverted U- shaped pattern of backward asking, the asking curves are steeper for the igraine groups with thresholds particularly elevated when the ask is ost disruptive (Figs. 3, 4). The tiing of asking effects has been used to infer the duration of cortical processing. In backward asking, if the target coputation is incoplete when the ask is presented, it can interrupt target processing and ipair perception. The teporal window over which a ask exerts its effect is thought to reflect the duration of the underlying coputation. 35,36,46 If igraineurs had a processing delay, the asking functions should be shifted horizontally and the peak asking effect would be seen for longer SOAs. Such a shift clearly is not supported by the data, and the peak SOA is the sae for controls and igraineurs. This argues against a general processing delay in igraine. One of the ost consistent findings in igraine is a lack of habituation for sensory evoked potentials. 47 Prolonged expo-

7 1166 Wagner et al. IOVS, February 2013, Vol. 54, No. 2 sure to a high-contrast reversing checkerboard results in a reducing aplitude for pattern-reversal visual evoked potential (VEP) in noral observers. The VEP aplitude for igraineurs by contrast does not reduce and in soe cases it ay even increase (potentiation). Siilarly, altered visual adaptation effects have been reported recently in igraine. The typical adaptation effect of decreased sensitivity following adaptation to a high contrast flicker was seen for norals and igraineurs when tested with low contrasts. 48 However, following adaption igraineurs showed the opposite effect for high contrast: increased contrast sensitivity. This facilitation effect was not observed in controls. Psychophysical studies of visual aftereffects also have described longer aftereffect durations in igraineurs. 12 These findings likewise point toward altered cortical adaptation processes. It ay be that the sae processes underlie the deficiencies easured psychophysically and with VEPs. 48 Repeated exposure to the shapes in our experient is inevitable and the possibility of this resulting in adaptation effects ust be considered. Individuals ran different conditions (with and without ask as well as different SOAs) in different blocks. The order in which SOAs were run was randoized across participants. Regular breaks were taken between blocks, reducing the possibility of significant adaptation effects across conditions. It nevertheless reains a possibility that an adaptation effect ay have occurred during blocks. Given the balanced design of the order in which different SOA conditions were copleted, we think it unlikely that the pattern of results with increased thresholds ainly for interediate SOAs in igraine is the result of unifor adaptation. In the absence of a specific effect of adaptation for soe but not other SOAs, we would expect all SOAs to be affected siilarly, resulting in asking functions that are elevated equally throughout the range of SOAs. This is not what we observed. That said, the PTM, on which our siulations were based, was developed originally to describe contrast adaptation effects. 39 Hence, our results, as well as those described above ay point to a coon deficit in igraine related to the process of adaptation/habituation. Based on the potentiation observed with high contrasts when easuring contrast sensitivity following adaptation, one ight have predicted an enhanced perforance for the igraine groups with high contrast stiuli in our experients. The asking effect in our case, however, is ore likely to arise as the result of interactions between shape-specific echaniss at higher stages of visual processing, rather than interactions at the relatively early levels underpinning contrast adaptation effects. Future investigations are required to address this issue. Masking has been attributed to cortical inhibition between neurons tuned to target and asks. 49 If asking is a consequence of inhibitory interactions, decreased inhibition in igraineurs should result in less asking copared to controls. It has been shown recently that the better perforance of igraineurs on a etacontrast asking task can be explained by a generally higher sensitivity copared to controls rather than a difference in asking. 9 Our results are consonant with the notion that inhibition generally is not reduced in igraine, at least for those neuronal pathways that are responsible for asking. 9,50 An alternative account for the differences between igraineurs and controls is neuronal noise. An increase in internal noise can explain why igraineurs perfor poorer than controls in a range of tasks. 11,18,19,22,51,52 Indeed, those studies explicitly deterining the aount of internal noise 18,19,24 have found it to be abnoral in igraine. To provide an analysis of the effect of increased internal noise levels in our study, we applied a odel that has been used widely in estiating internal noise, 37 which recently has been odified to include a divisive gain-control process that can be applied to asking 39 (Appendix). The odel has a nuber of paraeters, including two internal noise sources occurring before (N 1 ) and after (N 2 ) the contrast gain control, siilar to additive and ultiplicative noises in the original odel. 39 Increasing the early noise (N 1 ) akes incorrect predictions (Fig. 5). Increasing the late noise (N 2 ) predicts a general upward shift of the asking function that captures the higher thresholds in igraine, but would overestiate the difference between controls and igraine for short and long SOAs. Hence, neither increased additive nor ultiplicative noise is sufficient to explain the pattern of reduced sensitivity in igraine. Another explanation is required. The only factor that produces a steeper asking function is the nonlinearity within the neuronal network responsible for asking. This is reasonable, since the transducer function is regarded as a key coponent for odels explaining pattern asking. 53,54 An abnorally high nonlinearity in the gain-control process coupled with an increased ultiplicative internal noise in igraine can capture the data successfully with and without asking (Fig. 6). Both factors require a higher increase for igraineurs who experience visual auras (MA) copared to those who do not (MO). Acknowledgents The authors thank Gordon Dutton and Peter Storch for their help with various parts of this study. Peter Storch allowed us access to igraine patients through the Middle-Geran Headache Centre, Neurology Departent, University Clinic Jena, Jena, Gerany. APPENDIX Gain-Control Model for Masking The odel is a odified version of the PTM 37,38 that has been used widely to quantify the aount of various internal noise sources. In the original odel, the signal initially is filtered (perceptual teplate) and the filter response is passed through a nonlinear transducer function, consistent with psychophysical observations (e.g., Weber s law in perceptual tasks 53 ). This is subject to two internal noise sources: ultiplicative noise, with variance that is proportional to the response agnitude elicited by the stiulus as well as additive noise with fixed, stiulus/response-independent variance. The final part consists of a decision stage. This odel can be applied directly to experiental conditions where external noise is added to the stiulus. Recently, a odified version of this odel, replacing the ultiplicative noise stage with a divisive contrast gain-control process, has been applied successfully to predict the detriental effect of visual adaptation on the visibility of sine-wave gratings. 39 Figure 5A illustrates the odel in the case of a target and a ask. An inhibitory gain control reduces the response of the target, consistent with physiological, psychophysical, and odeling evidence, which has shown that such an operation odifies the stiulus gain and allows neurons to reain in an optial regie despite large variation in stiulus intensities Contrast-gain control has been linked to the effect of asking on target visibility 39,40,43 and we have applied this odel to our data. Note that the gain control in this odel is related to the interactions between shape encoding processes unlike the interactions underlying contrast-gain control at the earlier stages of visual processing. The target and the ask are processed by perceptual teplates with gains of b and b, respectively. Consistent with psychophysical and iaging studies, the output of RF teplates should depend on a nuber of factors, including

8 IOVS, February 2013, Vol. 54, No. 2 Global Shape Processing Deficits in Migraine 1167 RF contour contrast, 28 aplitude, 26,30,36,55 radial frequency, 26,28,55 size, 56 and orientation. 36 We only consider contrast (c and c ) and aplitude (A and A ) here as the other paraeters did not vary in the experients: S ¼ bca S ¼ b c A These outputs for target and ask are subject to a nonlinearity (c and c ) odeled by a power function 53,54 : S 0 ¼ b c c c A c S 0 ¼ bc c c A c An internal noise is added after the transducer, but before the gain control with a ean of 0 and a standard deviation of N 1. The cobined activity (E) in the gain control pathway at this stage is: E ¼ S 02 þ S 02 þ N2 1 Divisive, inhibitory gain control reduces the response of the target (S 00 ): S 00 ¼ pffiffiffiffiffiffiffiffiffiffiffiffi ð4þ b þ E where b is a threshold to avoid the divisive gain ter taking on zero values if target and ask contrasts are close to zero. The noise variance siilarly is noralized by the sae divisive ter: r 02 ¼ N2 1 ð5þ b þ E A second internal noise source follows after the gain control, with a ean of 0 and standard deviation of N 2. The overall noise variance is given by the su of all noise variances: r 2 total ¼ N2 1 b þ E þ N2 2 ð6þ Note that the two noise sources in this odel, N 1 and N 2, are related to the activity of putative shape-encoding echaniss and are not luinance-related noise as in the original odel. 38 Finally, the stiulus discriinability is given by the total signal-to-noise ratio: d 0 ¼ S00 b c c c A c qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi r total N 2 2 ðb2c c 2c A 2c þ b 2c S0 2 c2c A 2c ð1þ ð2þ ð3þ ÞþN1 2 þ N2 2 ðb þ N2 1 Þ ð7þ The noinator describes the response to the stiulus and the denoinator describes the gain control, consisting of the su of the responses to the stiulus and the ask plus noise. Note that the two noise sources occurring before or after the gain control (N 1 and N 2 ) can be linked to additive and ultiplicative internal noises. 39 The target and ask contrasts were set to 1 for our experients in all asking conditions and c was set to zero in the baseline condition without ask. c and c describe the behavior of the nonlinear transducers, and are assigned different paraeters to allow the gain control nonlinearity (c ) to behave different fro that responding to the target. Threshold aplitudes (A) for the shape discriination task for a given sensitivity d can be calculated by solving Equation 7 for A: 2 A s ¼ 4 N2 2 b2c c2c 3 þ N1 2 þ N2 2 ðb þ N2 1 Þ 5 A 2c b 2c d 0 N2 2 b2c To produce the typical U-shaped backward asking profile, we assue that the gain of the ask (b ) is a function of tie (SOA). The odel does not include the teporal dynaics that underlie backward asking. A unified odel for backward asking that could be applied to the data is not available. Several odels have been proposed (see the study of Francis 57 for a suary) that can capture soe, but not all aspects of asking. 57 While all odels produce the typical U-shaped asking function and are siilar in spirit, 57 they differ substantially in detail. Therefore, we assued a asking effect with agnitudes depending on SOA and restricted our analysis to the effects of a nuber of paraeters, which have been considered in earlier studies on igraine, 18,19 including internal noise, processing efficiency, and response nonlinearity. In addition to the dependence on SOA, asking for contour discriination also has been shown to depend on a nuber of static paraeters, including the ask aplitude, phase relative to the target as well as ask size. 36 Soe of these effects are highly nonlinear. Such individual effects could be included into the odel by introducing separate variables, but given that they were not varied in our experients and because our ai was to copare perforance between groups, they were collapsed into a single variable, A. The baseline threshold aplitude (A s,base ) for shape discriination without a ask is given by setting c ¼ 0in Equation 8: " #1 A s;base ¼ N2 1 þ N2 2 ðb þ N2 1 Þ 2c ð9þ b 2c d N b2c References 1 2c ð8þ 1. Wilkinson F. Auras and other hallucinations: windows on the visual brain. Prog Brain Res. 2004;144: Lauritzen M. Pathophysiology of the igraine aura the spreading depression theory. Brain. 1994;117: Welch KMA, Dandrea G, Tepley N, Barkley G, Raadan NM. The concept of igraine as a state of central neuronal hyperexcitability. Neurol Clin. 1990;8: Wilkins A, Niosith I, Tait A, et al. A neurological basis for visual discofort. Brain. 1984;107: Chronicle E, Mulleners W. Might igraine daage the brain. Cephalalgia. 1994;14: Paler JE, Chronicle EP, Rolan P, Mulleners WM. Cortical hyperexcitability is cortical under-inhibition: evidence fro a novel functional test of igraine patients. Cephalalgia. 2000; 20: Alpern M. Metacontrast. J Opt Soc Aer. 1953;43: Breiteyer BG, Ganz L. Iplications of sustained and transient channels for theories of visual-pattern asking, saccadic suppression, and inforation-processing. Psychol Rev. 1976; 83: Shepherd AJ, Wyatt G, Tibber MS. Visual etacontrast asking in igraine. Cephalalgia. 2011;31: McColl SL, Wilkinson F. Visual contrast gain control in igraine: easures of visual cortical excitability and inhibition. Cephalalgia. 2000;20: Wilkinson F, Crotogino J. Orientation discriination thresholds in igraine: a easure of visual cortical inhibition. Cephalalgia. 2000;20:57 66.

9 1168 Wagner et al. IOVS, February 2013, Vol. 54, No Shepherd AJ. Increased visual after-effects following pattern adaptation in igraine: a lack of intracortical excitation? Brain. 2001;124: Shepherd AJ, Paler JE, Davis G. Increased visual after-effects in igraine following pattern adaptation extend to siultaneous tilt illusion. Spat Vis. 2002;16: McKendrick AM, Badcock DR. Motion processing deficits in igraine. Cephalalgia. 2004;24: Shepherd AJ. Visual contrast processing in igraine. Cephalalgia. 2000;20: McKendrick AM, Cioffi GA, Johnson CA. Short-wavelength sensitivity deficits in patients with igraine. Arch Ophthalol. 2002;120: Shepherd AJ. Color vision but not visual attention is altered in igraine. Headache. 2006;46: Wagner D, Manahilov V, Loffler G, Gordon GE, Dutton GN. Visual noise selectively degrades vision in igraine. Invest Ophthalol Vis Sci. 2010;51: Webster KE, Dickinson JE, Battista J, McKendrick AM, Badcock DR. Increased internal noise cannot account for otion coherence processing deficits in igraine. Cephalalgia. 2011;31: Wagner D, Manahilov V, Gordon G, Storch P. Long-range inhibitory echaniss in the visual syste are ipaired in igraine sufferers. Cephalalgia. 2012;32: McKendrick AM, Badcock DR, Gurgone M. Vernier acuity is noral in igraine, whereas global for and global otion perception are not. Invest Ophthalol Vis Sci. 2006;47: Ditchfield JA, McKendrick AM, Badcock DR. Processing of global for and otion in igraineurs. Vision Res. 2006;46: Antal A, Tee J, Nitsche MA, Varga ET, Lang N, Paulus W. Altered otion perception in igraineurs: evidence for interictal cortical hyperexcitability. Cephalalgia. 2005;25: Webster KE, Dickinson JE, Battista J, McKendrick AM, Badcock DR. Evidence for increased internal noise in igraineurs for contrast and shape processing. Cephalalgia. 2012;32: Wilkinson F, Wilson HR, Habak C. Detection and recognition of radial frequency patterns. Vision Res. 1998;38: Loffler G, Wilson HR, Wilkinson F. Local and global contributions to shape discriination. Vision Res. 2003;43: Hess RF, Wang YZ, Dakin SC. Are judgeents of circularity local or global? Vision Res. 1999;39: Schidtann G, Kennedy GJ, Orbach HS, Loffler G. Non-linear global pooling in the discriination of circular and noncircular shapes. Vision Res. 2012;62: Pasupathy A, Connor CE. Population coding of shape in area V4. Nat Neurosci. 2002;5: Wilkinson F, Jaes TW, Wilson HR, Gati JS, Menon RS, Goodale MA. An fmri study of the selective activation of huan extrastriate for vision areas by radial and concentric gratings. Curr Biol. 2000;10: Loffler G. Perception of contours and shapes: Low and interediate stage echaniss. Vision Res. 2008;48: Headache Classification Subcoittee of the International Headache Society: The International Classification of Headache Disorders: 2nd Edition. Cephalalgia. 2004;24(suppl 1): Lipton RB, Stewart WF, Sawyer J, Edeads JG. Clinical utility of an instruent assessing igraine disability: The Migraine Disability Assessent (MIDAS) questionnaire. Headache. 2001;41: Pelli DG, Zhang L. Accurate control of contrast on icrocoputer displays. Vision Res. 1991;31: Breiteyer B. Visual Masking: An Integrative Approach. New York: Oxford University Press; Habak C, Wilkinson F, Wilson HR. Dynaics of shape interaction in huan vision. Vision Res. 2006;46: Lu ZL, Dosher BA. Characterizing observers using external noise and observer odels: assessing internal representations with external noise. Psychol Rev. 2008;115: Lu ZL, Dosher BA. Characterizing huan perceptual inefficiencies with equivalent internal noise. J Opt Soc A A Opt Iage Sci Vis. 1999;16: Dao DY, Lu ZL, Dosher BA. Adaptation to sine-wave gratings selectively reduces the contrast gain of the adapted stiuli. J Vis. 2006;6: Wilson HR, Huanski R. Spatial frequency adaptation and contrast gain control. Vision Res. 1993;33: Heeger DJ. Noralization of cell responses in cat striate cortex. Visual Neuroscience. 1992;9: Sclar G, Lennie P, Depriest DD. Contrast adaptation in striate cortex of acaque. Vision Res. 1989;29: Foley JM. Huan luinance pattern-vision echaniss: asking experients required a new odel. J Opt Soc A A Opt Iage Sci Vis. 1994;11: Anderson ND, Habak C, Wilkinson F, Wilson HR. Evaluating shape after-effects with radial frequency patterns. Vision Res. 2007;47: McKendrick AM, Badcock DR, Badcock JC, Gurgone M. Motion perception in igraineurs: abnoralities are not related to attention. Cephalalgia. 2006;26: Loffler G, Gordon GE, Wilkinson F, Goren D, Wilson HR. Configural asking of faces: evidence for high-level interactions in face perception. Vision Res. 2005;45: Abrosini A, Schoenen J. Electrophysiological response patterns of priary sensory cortices in igraine. J Headache Pain. 2006;7: Karanovic O, Thabet M, Wilson HR, Wilkinson F. Detection and discriination of flicker contrast in igraine. Cephalalgia. 2011;31: Weisstein N, Ozog G, Szoc R. Coparison and elaboration of 2 odels of etacontrast. Psychol Rev. 1975;82: Huang J, DeLano M, Cao Y. Visual cortical inhibitory function in igraine is not generally ipaired: evidence fro a cobined psychophysical test with an fmri study. Cephalalgia. 2006;26: McKendrick AM, Johnson CA, Anderson AJ, Fortune B. Elevated vernier acuity thresholds in glaucoa. Invest Ophthalol Vis Sci. 2002;43: McKendrick AM, Badcock DR. An analysis of the factors associated with visual field deficits easured with flickering stiuli in-between igraine. Cephalalgia. 2004;24: Nachias J, Sansbury RV. Grating contrast discriination ay be better than detection. Vision Res. 1974;14: Foley JM, Legge GE. Contrast detection and near-threshold discriination in huan-vision. Vision Res. 1981;21: Bell J, Wilkinson F, Wilson HR, Loffler G, Badcock DR. Radial frequency adaptation reveals interacting contour shape channels. Vision Res. 2009;49: Mullen KT, Beaudot WHA, Ivanov IV. Evidence that global processing does not liit thresholds for RF shape discriination. JVis. 2011; Francis G. Quantitative theories of etacontrast asking. Psychol Rev. 2000;107: