Available online at www.sciencedirect.com

03 (2008) 275–282www.elsevier.com/locate/schres

Schizophrenia Research 1

Altered ‘three-flash’ illusion in response to two light pulsesin schizophrenia

Dan Norton a, Dost Ongur a,b, Charles Stromeyer III c, Yue Chen a,b,⁎

a McLean Hospital, United Statesb Harvard Medical School, United States

c Harvard University, United States

Received 26 November 2007; received in revised form 27 February 2008; accepted 5 March 2008Available online 18 April 2008

Abstract

Background: Disorganization is a core dysfunction in schizophrenia. Coherent thought and behavior rely on the interactive neuralresponses to temporally discrete external events. Previous studies have demonstrated that a single visual stimulus (event) isabnormally affected by another (as in backward masking), but the integration (or ‘synthesis’) of temporally discrete events remainslargely unexplored in schizophrenia. We examined the perceived interaction of two elementary visual events in schizophreniapatients, using a psychophysical approach.Methods: Two brief, spatially-coincident foveal light pulses (5 ms) were presented with different inter-stimulus intervals (ISIs). AtISIs around 100 ms, a substantial proportion of the light pulse pairs was paradoxically perceived as three flashes, a knownphenomenon in normal subjects. The subjects reported the number of flashes perceived (‘one’, ‘two’ or ‘three’).Results: Schizophrenia patients (n=28) reported fewer ‘three flashes’ than normal controls (n=26) at the ISIs where ‘three flash’reports were greatest in normal subjects (90 to 110 ms). On the other hand, at longer ISIs (130–310 ms) patients reported ‘threeflashes’ in more trials than did normal subjects. The perception of three flashes in patients was correlated with certain aspects ofclinical status, including the positive and general subscales of the PANSS.Discussion: The alteration of the ‘three-flash’ illusion in schizophrenia suggests that the synthesis of discrete visual events istemporally ‘dilated’ or distorted, which might contribute to disorganized thought and behavior.© 2008 Elsevier B.V. All rights reserved.

Keywords: Visual processing; Schizophrenic; Temporal organization; Sensory; Cognitive

1. Introduction

Disorganized thought and behavior are prevalent inschizophrenia. For normal behavior, one must respond toa host of temporally discrete stimuli and integrate, or

⁎ Corresponding author. Room G06, Centre Bldg., 115 Mill Street,Belmont, MA 02478, United States. Tel.: +1 617 855 3615.

E-mail address: ychen@mclean.harvard.edu (Y. Chen).

0920-9964/$ - see front matter © 2008 Elsevier B.V. All rights reserved.doi:10.1016/j.schres.2008.03.002

‘synthesize’, the responses to these stimuli. Alteration ofthis synthesis might cause a coherent flow of events tobecome a barrage of more disjointed information. Study-ing the synthesis of temporally disparate events may thusshed light on the pathophysiology of schizophrenia.

Altered temporal processing of visual information inschizophrenia has been shown in studies on visual back-ward masking (Green et al., 1994a; Butler et al., 1996) andtemporal contrast detection (Butler et al., 2001a; Slaghuis,

276 D. Norton et al. / Schizophrenia Research 103 (2008) 275–282

1998; Chen et al., 2000; Keri et al., 2000). These studiesreveal two general effects in schizophrenia. First, there is areduced sensitivity for contrast detection of temporallymodulated stimuli. Second, there is temporal elongation inbackward masking—a measure of how one visual event isaffected by a later, masking event. The present studyevaluated another type of temporal interaction in schizo-phrenia—the way in which two equivalent temporallydiscrete events interact to yield a ‘synthesized’ perceptualresponse.

We measured the synthesized perceptual response totwo temporally discrete visual events (brief light pulses)in both patients and normal controls, and we compared

Fig. 1. Model of the neural response to single light flashes (top andmiddle panelshorizontal dashed lines represent the hypothetical threshold that must be exceeresponse to a single light pulse presented at time t0. The middle left panel shows~100ms after t0). The bottom left panel is the summed responses for the pair of psame schema, but the second flash occurs at t2, about 100 ms after t0, leadingperceived.

this to the processing of single visual events (temporalcontrast detection). The results were evaluated in termsof relevant clinical features.

We adopted the paradigm of Bowen (1989). Twosupra-threshold, spatially-coincident light pulses (5 ms)were presented to the fovea with varying inter-stimulusintervals (ISIs). For ISIs around 100ms the two pulses areregularly mis-perceived as three flashes by normalsubjects (Bowen, 1989). At this ISI, the ‘three-flash’illusion occurs for 40–70% of the presentations. Accord-ing to Bowen, the ‘three-flash’ illusion represents theinteraction (the linear sum) of the multi-phase ImpulseResponse Functions (IRF) to each of the two flashes. The

) and their linear sum (bottompanels)—adopted fromBowen (1989). Theded for perceiving a flash—as in Bowen. a. The top left panel shows thethe neural response to the second pulse, presented at time t1, (greater thanulses—at this ISI, two flashes are perceived. b. The right panels show theto a peak to peak summation of the two IRFs, —now three flashes are

Table 1Demographic information of the sample

Group Demographic variable

Sex Age VerbalIQ⁎

Education⁎ Parentaleducation

Normal controls(NC)

F=16 43.0 108.0 15.9 13.5M=10 (13.8) (9.2) (2.3) (2.6)

Schizophreniapatients (SZ)

F=17 40.3 100.8 14.1 13.9M=11 (9.9) (12.3) (2.0) (2.1)

Means are reported above standard deviations. Education and age arein years. Verbal IQ was measured using the WAIS-R (Wechsler, 1981).Asterisk denotes significant group difference ( pb0.05).

277D. Norton et al. / Schizophrenia Research 103 (2008) 275–282

multi-phase IRFs have at least three alternating (or‘oscillating’) lobes of excitation and inhibition thatallow the perception of a third flash from only two pulsestimuli (see Fig. 1 as well as Discussion for furtherexplanation). These IRFs and their summation likelyoccur in early stages of the magnocellular pathway—apathway largely responsible for temporal processing andmotion detection (Bowen, 1989; Stromeyer and Martini,2003; Stromeyer et al., 1997).

2. Methods

2.1. Subjects

Subjects were 28 patients with schizophrenia orschizoaffective disorder and 26 non-psychiatric controls.The patients met the criteria for schizophrenia orschizoaffective disorders based on standardized interviewsby independent clinicians using the SCID-IV (First et al.,2002a) and all available medical records. The schizo-phrenia group comprised inpatients from the Schizophre-nia and Bipolar Disorder Program at McLean Hospital(n=6) and outpatients treated at McLean Hospital or localclinics (n=22). All were taking antipsychotic medication(mean chlorpromazine (CPZ) equivalent was 537.2 mg(SD=358.1 mg)) (Woods, 2003; Davis, 1974). Fifteenpatients were taking antidepressant medication, 7 weretaking anti-anxiety medications, and 17 were taking moodstabilizers. The patient group hadmean PANSS (Kay et al.,1987) scores of 15.8 (SD=6.8), 13.6 (SD=5.9) and 29.4(SD=10.6) for the positive, negative and general scales,respectively. Average duration of illness was 18.0 years(SD=10.4 years). Normal healthy controls were screenedfor axis-one disorders using an interview based on theSCID N/C (First et al., 2002b). Subjects from both groupswere excluded if they had history of brain injury orneurological disorders, a verbal IQ (Wechsler, 1981) lowerthan 70, or any drug or alcohol abuse in the 6 months priorto the study. All subjects signed informed consent inaccordance with McLean Hospital's IRB guidelines.Clinical and demographic information are provided inTable 1. Patients and controls did not differ in age or sex,the standard deviation for age in the patient sample,however, was smaller than for controls. The two groupsdiffered in education (t=2.4, p=0.02) and IQ (t=2.3,p=0.02).

2.2. Stimulus and procedure

2.2.1. Perception of light pulse pairsA diffuse red light emitting diode (LED) provided

light pulses. Subjects viewed the LED target in a

weakly-illuminated room—fixation was guided by thecenter of a small, circular neutral density filter placedover the LED. The target was a nearly uniform disc,subtending 17 min arc, with a peak luminance of 35 cd/m2. The LED has extremely high temporal bandwidthand provides precisely controlled flashes. Such flashesallow the visual events to be manipulated solely in thetemporal domain. A Macintosh G3 computer controlledthe presentation of the light pulses. For each trial, thelight pulse was presented twice, each for 5 ms, as inBowen's paradigm. The inter-stimulus interval (ISI) ofthe two pulses ranged over 10 values (30, 70, 90, 110,130, 150, 190, 230, 270 or 310 ms). Each ISI had 10trials. A total of 100 trials were presented in a quasi-random order across ISIs.

On each trial the subject was informed that they wouldsee one, two, or three flashes, and was asked to report thenumber of flashes they saw. The perceived number offlashes was the dependent variable. Education and IQscores, the only two demographic variables that differedbetween groups, were used as covariates in analysis.

2.2.2. Contrast detectionTemporal contrast detection was also measured to

provide an index for processing of a single visualstimulus. The target was a sinusoidal grating (0.5 cyclesper degree) moving left or right at a temporal frequency of0, 3, 6 or 12 Hz (or at a speed of 0, 6, 12 or 24 deg/s). Thestimulus field subtended 13 deg, at a mean luminance of35 cd/m2. The threshold was measured separately at eachtemporal frequency, using a two alternative forced-choicestaircase to determine the minimum contrast value for79% accuracy. Each trial had two temporal intervals forthe target and a blank respectively, ordered randomly, andthe task was to judge which of the two intervals containedthe target. The target and blank each lasted 300 ms,separated by a 500 ms gap. Contrast detection thresholdswere converted to contrast sensitivity (1/threshold).

Fig. 3. Contrast sensitivity (inverse of contrast detection threshold) as a

278 D. Norton et al. / Schizophrenia Research 103 (2008) 275–282

3. Results

3.1. Perception of the light pulse pair

Perception of the ‘three-flash’ illusion occurred forschizophrenia patients (SZ) less frequently at short ISIs(90 to110 ms) and more frequently at intermediate andlong ISIs (130 to 310 ms) (Fig. 2). A two way ANOVAwas performed on the proportion of ‘three-flash’ trials,using diagnosis and ISI as factors. The diagnosis X ISIinteraction was significant (F=2.94, p=0.002), indicatingthat the ‘three-flash’ illusion for SZ and NC peakedat different ISIs (Fig. 2a). There was a main effect for

Fig. 2. Perception of pulse pairs. a. The proportion of trials perceived as‘three flashes’ at varying inter-stimulus intervals ISIs. b. The proportion oftrials perceived as ‘two flashes’, and c. the proportion of trials perceived as‘one flash’. In all panels, error bars represent ±1 standard error.

function of temporal frequency. The visual target was a sinusoidal grating(0.5 cycles per degree) moving at different velocities (0 to 24 deg/s), orequivalently modulating at different temporal frequencies (0 to 12 Hz).The error bars represent ±1 standard error.

ISI (F=13.97, pb0.001), but not for diagnosis (F=0.04,p=0.85).

For the proportion of trials perceived as ‘one flash’,ANOVAwith diagnosis and ISI as factors showed maineffects for ISI (F=277.44, pb0.001) and diagnosis(F=8.01, p=0.005). There was also a trend towards adiagnosis X ISI interaction (F=1.80, p=0.07) (Fig. 2b).

For the proportion of trials perceived as ‘two flashes’,there were significant effects for ISI (F=122.3, pb0.001)and group (F=6.4, p=0.012) (Fig. 2c). The interactionwas insignificant (F=1.2, p=0.29).

3.1.1. Contrast detectionContrast sensitivities for SZ were lower than for NC

across all temporal frequencies (0, 3, 6 and 12Hz, Fig. 3)—the group difference reached a statistically significant level(F=18.1, pb0.001). There was also a main effect fortemporal frequency (F=36.2, pb0.001), but the interactionbetween group and temporal frequency was not significant(F=1.6, p=0.18).

3.1.2. Relationship between ‘three-flash’ illusion andother variables

3.1.2.1. Visual variables. Table 2 shows correlationcoefficients between the ‘three-flash’ illusion at peak

Table 2Correlations between perception of ‘three flashes’ and contrast sensitivity

Group (ISI) Contrast detection temporal frequency

0 Hz 3 Hz 6 Hz 12 Hz

SZ (90 ms) −0.15 −0.22 0.13 0.30SZ (130 ms) −0.15 −0.02 0.01 0.05NC (90 ms) 0.03 −0.08 0.36 0.38NC (130 ms) −0.11 −0.08 0.34 0.37

Table 3Correlations between demographic variables and the perception of the‘three flashes’

Group (ISI) Demographic variable

Verbal IQ Education Parental education Age

SZ (90 ms) −0.14 0.05 0.13 −0.51⁎SZ (130 ms) −0.29 −0.23 0.05 −0.41⁎NC (90 ms) −0.03 −0.30 0.14 −0.22NC (130 ms) 0.13 −0.16 0.27 −0.30

Asterisk denotes pb0.05.

279D. Norton et al. / Schizophrenia Research 103 (2008) 275–282

ISIs (90 ms for NC, and 130 ms for SZ) and contrastsensitivity in patients and controls. Although none ofthe correlations were significant, normal controls ap-proached significance for temporal frequencies of 6(r=0.36, p=0.08) and 12 Hz (r=0.38, p=0.07).Patients showed such a pattern at 12 Hz (r=0.30,p=0.14).

3.1.2.2. Demographic variables. Table 3 showscorrelation coefficients between the ‘three-flash’ illu-sion and several demographic variables. In normalcontrols, the prevalence of the ‘three-flash’ illusion wasnot significantly correlated with IQ, education, parentaleducation or age. In patients, the ‘three-flash’ illusionwas correlated with age (r=−0.51, pb0.01 at 90 ms, r=−0.41, pb0.05 at 130 ms), but not with IQ, education orparental education.

3.1.2.3. Clinical variables. The prevalence of the‘three-flash’ illusion in patients at 130 ms was correlatedwith positive (r=0.65, pb0.001) and general (r=0.64,pb0.001) subscales of the PANSS. To a lesser extent,the negative subscale of the PANSS was also correlatedwith the ‘three-flash’ illusion measure at 130 ms(r=0.46, pb0.05). To evaluate the relationship betweenthe ‘three-flash’ illusion and clinical disorganization, weused a cluster score composed of three PANSS items(Conceptual Disorganization, Difficulty in AbstractThinking and Disorientation (Lepine et al., 1989)). Thecluster score was correlated with the perceptual measureat the 130 ms ISI (r=0.58, pb0.01) but not at the 90 msone (r=0.25, pN0.10).

Neither CPZ equivalents (r=0.17) nor illnessduration (r=−0.18) were correlated with the ‘three-flash’ illusion at 130 ms. We also compared the three-flash illusion in the presence versus the absence of otherdrugs (antidepressant, anti-anxiety and mood stabiliz-ing medications) across ISIs; each drug by a separateANOVA. No medication group effects or interactionswere significant ( pN0.05).

4. Discussion

We observed that the perceptual response to twosimple visual events (light pulses) was altered in schizo-phrenia. The perception of the ‘three-flash’ illusion, typi-cally occurring for ISI's near 100 ms in healthy subjects,shifted to longer intervals in patients. At longer intervals(beyond 110 ms), patients perceived three flashes morefrequently, and two flashes less frequently. Also, patientsperceived one flash more frequently than controls atintervals from 90 to 130 ms. This pattern of results sug-gests that the underlying temporal impulse responsefunction is prolonged (or ‘dilated’) in schizophrenia.

4.1. Temporal interactions

The temporal interactions that give rise to the ‘three-flash’ illusion likely begin at the magnocellular ganglioncells of the retina and lateral geniculate nucleus of thethalamus (LGN) (Stromeyer andMartini, 2003; Stromeyeret al., 1997). The magnocellular pathway largely mediatestemporal and motion processing, and this pathway isaffected by contrast gain control (Stromeyer and Martini,2003; Stromeyer et al., 1997). Contrast gain control ismanifested as a speeding-up of the IRF (and temporalprocessing) as stimulus contrast is raised in the magnocel-lular system [reviewed in 8].

The perception of the ‘three-flash’ illusion (with thedouble-pulse stimulus) requires that the IRF has at leastthree phases of alternating excitation and inhibition—modeled by a dampened sine function (Bowen, 1989)(Fig. 1). The perceived number of flashes is predicted by thesum of a tri-phasic (Bowen, 1989) IRF to each of the twolight pulses. This sum depends on the magnitude of eachIRF (controlled by flash intensity) and the delay betweenthe two IRFs (controlled by ISI). The perception of the thirdflash arises (at the appropriate ISI) by a linear summationbetween the third positive, excitatory lobe of the first IRF(from the first pulse) and the second excitatory lobe of thesecond IRF (from the second pulse). It is the ‘ringing’ of theextra lobes that results in the perception of the extra flash.

In schizophrenia the ‘three-flash’ illusion (Fig. 2a) wasreduced at briefer ISIs and increased at longer ISIs(compared to normal subjects). This shift of the illusion tolonger intervals could be explained by a temporal dilationof the IRF in schizophrenia. Such dilationwould cause thesummation of the second and high-order phases of theresponses to be delayed accordingly, causing the ‘three-flash’ illusion to shift to longer ISIs.

The prolongation of the IRF in schizophrenia is con-sistent with earlier results on backward masking (Greenet al., 1994b; Butler et al., 2001b), for the patients there also

280 D. Norton et al. / Schizophrenia Research 103 (2008) 275–282

required a longer temporal separation between mask andtarget to elicit optimal backward masking. The maindifference between the two types of studies is that thebackward masking studies examined the effect of a largemask on a subsequent different stimulus whereas thepresent study examined simple interactions of two virtuallyidentical stimuli (equal pulses). The pulse interaction,occurring at the exactly same spatial location and with nocognitive demands (such as object recognition), may bemuch simpler to interpret, as shown by the simple linearsummation model in Fig. 1.

4.2. Temporal interaction and processing of individualstimuli

In the present study, temporal contrast sensitivity wasmeasured as an index of visual processing of individualstimuli. Temporal contrast sensitivity was reduced inschizophrenia patients (Fig. 3). The reduced contrastsensitivity might indicate that contrast gain control isreduced in the magnocellular pathway for these patients,and this might lead to dilated IRFs (since contrast gain actsto speed-up the IRF (Stromeyer et al., 1997)). Reducedcontrast sensitivity at low temporal frequency could beconsistent with the dysfunction in the parvocellularpathway. Yet neither group showed any correlation be-tween the ‘three-flash’ illusion and contrast sensitivity forthe static gratings (0 Hz) (Table 2). On the other hand, thethree-flash illusion was modestly correlated contrastsensitivity at intermediate and high temporal frequenciesin controls, and to a lesser degree in patients. This mayhint at an association of the altered temporal interactionwith the functioning of the magnocellular pathway, butless so with the parvocellular pathway (Watson et al.,1977; Chen et al, 1996). However, the relatively smallmagnitudes of the correlations between the ‘three-flash’illusion and contrast sensitivity suggest that the altered‘three-flash’ illusion cannot be solely attributed to theprocessing of individual stimuli within the magnocellularpathway.

The role of the magnocellular pathway in schizo-phrenia is complex (Skottun and Skoyles, 2007; Delordet al., 2006; Butler and Javitt, 2005). Adding to thiscomplexity are the inherent attentional demandsinvolved in behavioral assessment of visual functioning(Joseph et al., 1997). The impaired ability to attend tobehavioral tasks in schizophrenia patients (Nestor et al.,2007; Stip et al., 1999) could be a factor for the alteredtemporal interaction studied here. However, the ‘three-flash’ illusion in patients is not universally dampened oramplified across conditions, rather it is shifted to specifictemporal intervals. The temporal-interval-specific result

cannot be easily reconciled with the attention interpreta-tion, which would likely alter the ‘three-flash’ illusionacross temporal intervals.

4.3. Temporal interaction and clinical manifestations

Altered temporal interactions of simple visual responsesmay give rise to more complex disorganized behavior. Inthe present study, the ‘three-flash’ illusion was correlatedwith the positive and general subscales of the PANSS butwas less correlated with the negative subscale. Also, thecluster score used to measure clinical disorganization wassignificantly and positively correlated with the ‘three-flash’illusion at 130 ms, but not at 90 ms. This pattern ofcorrelations suggests that those patients whose ‘three-flash’illusion had migrated to longer stimulus intervals exhibitmore positive symptoms such as clinical disorganization.No causal link can be drawn from the correlations betweenaltered temporal interactions and clinical measures ofdisorganization. Yet, since the temporal processing in the‘three-flash’ illusion occurs early on in the visual stream, itis reasonable to speculate that altered temporal interactionat the sensory stage contributes to higher order cognitivedisorganization such as Difficulty in Abstract Thinking andConceptual Disorganization. It is worth noting that inprevious studies, backward masking performance was notcorrelated with clinical symptoms in patients (Green andWalker, 1986; Butler et al., 2003). Clearly, more studies areneeded to better determinewhat visual measures are relatedto disorganized symptomatology.

4.4. Further studies

One area for future studies should be the character-ization of physiological correlates of temporal responsefunctions and the interaction between them. While someevidence suggests neuronal number and volume are notaltered in early visual areas such as LGN (Selemon andBegovic, 2007), electrophysiological studies wouldprovide complementary information about neural orga-nization of temporally discrete events in schizophrenia.

Another unanswered question concerns how thealtered temporal processing in schizophrenia is relatedto other visual and cognitive processes requiringtemporal organization (O'Donnell et al., 1996; Chenet al., 2003; Kim et al., 2006; Keri et al., 2002; Amadoand Olie, 2006; Bedwell et al., 2003; Park and Holzman,1992). Many cognitive activities such as reading(Revheim et al., 2006) and language processing(Adams et al., 1993) depend on organization of eventsthat occur at different points of time. It is conceivablethat the sensory temporal interaction studied here may

281D. Norton et al. / Schizophrenia Research 103 (2008) 275–282

contribute to poor performance in such commonactivities.

4.5. Summary

Our study demonstrated that the synthesis of temporalinteraction is lengthened in schizophrenia, as measured bythe ‘three-flash illusion’. This perceptual abnormality maybe partially linked to the patients' decreased contrastsensitivity for temporalmodulation. The altered ‘three-flashillusion’ represents a genuine abnormality of temporalorganization at the basic sensory level. Future studies inschizophrenia may reveal whether these temporal altera-tions affect other cognitive activities requiring temporalprocessing.

Role of funding sourceFunding for this study was provided in part by an NIMH Grant

61824. The NIMH had no further role in study design; in thecollection, analysis and interpretation of data; in the writing of thereport; and in the decision to submit the paper for publication.

ContributorsDaniel Norton tested the subjects and undertook the data analysis.

He also wrote the first draft of the paper. Dost Ongur managed theclinical evaluations of patients and provided support for analyzingclinical aspects of the data. Charles Stromeyer III contributed to thedesign of the study and the writing of the paper. Yue Chen contributedto the design of the study and the writing of the paper. He oversaw theimplementation of the study. All authors contributed to and haveapproved the final manuscript.

Conflict of interestAll authors declare that they have no conflicts of interest.

AcknowledgementWe thank Ken Nakayama for help in the initial phase of the study.

We also thank Ryan McBain for comments on an early version of thepaper.

References

Adams, J., et al., 1993. ERP abnormalities during semantic processingin schizophrenia. Schizophr. Res. 10 (3), 247–257.

Amado, I., Olie, J.P., 2006. Effects of disorganization on choicereaction time and visual orientation in untreated schizophrenics.Dialogues Clin. Neurosci. 8 (1), 53–58.

Bedwell, J.S., Brown, J.M., Miller, L.S., 2003. The magnocellularvisual system and schizophrenia: what can the color red tell us?Schizophr. Res. 63 (3), 273–284.

Bowen, R.W., 1989. Two pulses seen as three flashes: a superpositionanalysis. Vis. Res. 29 (4), 409–417.

Butler, P.D., Javitt, D.C., 2005. Early-stage visual processing deficitsin schizophrenia. Curr. Opin. Psychiatry 18 (2), 151–157.

Butler, P.D., et al., 1996. Backward masking in schizophrenia:relationship to medication status, neuropsychological functioning,and dopamine metabolism. Biol. Psychiatry 40 (4), 295–298.

Butler, P.D., et al., 2001a. Dysfunction of early-stage visual processingin schizophrenia. Am. J. Psychiatry 158 (7), 1126–1133.

Butler, P.D., Schechter, I., Zemon, V., Schwartz, S.G., Greenstein, V.C.,Gordon, J., Schroeder, C.E., Javitt, D.C., 2001b. Dysfunction ofearly-stage visual processing in schizophrenia. Am. J. Psychiatry158 (7), 1126–1133.

Butler, P.D., Maddox, J., Herkavy-Friedman, J.M., Amador, X.F.,Raymond, R.G., Javitt, D.C., Gorman, J.M., 2003. Visual-backwardmasking deficits in schizophrenia: relationship to visual pathwayfunction and symptomatology. Schizophr. Res. 59, 199–209.

Chen, Y., Bedell, H.E., Frishman, L.J., 1996. Temporal-contrastdiscrimination and its neural correlates. Perception 25 (5), 505–522.

Chen, Y., et al., 2000. Contrast detection in schizophrenia. Arch. Gen.Psychiatry 57 (10), 995.

Chen, Y., et al., 2003. Processing of global, but not local, motiondirection is deficient in schizophrenia. Schizophr. Res. 61 (2-3),215–227.

Davis, J.M., 1974.Dose equivalence of the antipsychotic drugs. J. Psychiat.Res. 11, 65–69.

Delord, S., et al., 2006. Psychophysical assessment of magno- andparvocellular function in schizophrenia. Vis. Neurosci. 23 (3-4),645–650.

First, M.B., et al., 2002a. Structured clinical interview for DSM-IV-TRAxis I Disorders -Patient Edition (SCID - I/P, 11/2002 revision).Biometrics Research Department, New York, NY.

First, M.B., S.R., Gibbon, M., William, J.B.W., Structure ClinicalInterview for DSM -IV-TR Axis I Disorders - Non-patient Edition(SCID-I/NP, 11/2002b revision) New York, NY BiometricResearch Department, New York State Psychiatric Institute.

Green, M., Walker, E., 1986. Symptom correlates of vulernability tobackward masking in schizophrenia. Am. J. Psychiatry 143 (2),181–186.

Green, M.F., Nuechterlein, K.H., Mintz, J., 1994a. Backward maskingin schizophrenia and mania. II. Specifying the visual channels.Arch. Gen. Psychiatry 51, 945–951.

Green, M.F., Nuechterlein, K.H., Mintz, J., 1994b. Backward maskingin schizophrenia and mania. I. Specifying a mechanism. Arch.Gen. Psychiatry 51 (12), 939–944.

Joseph, J.S., Chun,M.M., Nakayama,K., 1997.Attentional requirementsin a ‘preattentive’ feature search task. Nature 387 (6635), 805–807.

Kay, S., Fiszbein,A.,Opler, L., 1987. The Positive andNegative SyndromeScale (PANSS) for schizophrenia. Schizophr. Bull. 13, 261–276.

Keri, S., et al., 2000. Contrast detection in schizophrenia. Arch. Gen.Psychiatry 57 (10), 995–996.

Keri, S., et al., 2002. Spatiotemporal visual processing in schizo-phrenia. J. Neuropsychiatry Clin. Neurosci. 14 (2), 190–196.

Kim, D., et al., 2006. Magnocellular contributions to impaired motionprocessing in schizophrenia. Schizophr. Res. 82 (1), 1–8.

Lepine, J.P., Piron, J.J., Chapotot, E., 1989. Factor analysis of thePANSS in schizophrenic patients. In: Stefanis, C.N., Soldatos, C.R,Rabavilas, A.D. (Eds.), Psychiatry Today: Accomplishments andPromises. Proceedings of the 8th International Congress ofPsychiatry, Athens: Amsterdam. Abstract 3232.

Nestor, P.G., et al., 2007. Attentional networks and cingulum bundle inchronic schizophrenia. Schizophr. Res. 90 (1-3), 308–315.

O'Donnell, B.F., et al., 1996. Selective deficits in visual perception andrecognition in schizophrenia. Am. J. Psychiatry 153 (5), 687–692.

Park, S., Holzman, P.S., 1992. Schizophrenics show spatial workingmemory deficits. Arch. Gen. Psychiatry 49 (12), 975–982.

Revheim, N., et al., 2006. Reading impairment and visual processingdeficits in schizophrenia. Schizophr. Res. 87 (1-3), 238–245.

Selemon, L.D., Begovic, A., 2007. Stereologic analysis of the lateralgeniculate nucleus of the thalamus in normal and schizophrenicsubjects. Psychiatry Res. 151 (1-2), 1–10.

282 D. Norton et al. / Schizophrenia Research 103 (2008) 275–282

Skottun, B.C., Skoyles, J.R., 2007. Contrast sensitivity and magnocel-lular functioning in schizophrenia. Vis. Res. 47 (23), 2923–2933.

Slaghuis, W.L., 1998. Contrast sensitivity for stationary and driftingspatial frequency gratings in positive- and negative-symptomschizophrenia. J. Abnorm. Psychology 107, 49–62.

Stip, E., et al., 1999. Atypical neuroleptics and selective attention.Encephale 25 (3), 260–264.

Stromeyer, C.F., Martini, P., 2003. Human temporal impulse responsespeeds up with increased stimulus contrast. Vis. Res. 43 (2003),285–298.

Stromeyer, C.F., et al., 1997. Colour adaptationmodifies the long-waveversus middle-wave cone weights and temporal phases in humanluminance (but not red-green) mechanism. J. Physiol. 499 (Pt 1),227–254 (Pt 1).

Watson, A.B., et al., 1977. Patterns of temporal interaction in thedetection of gratings. Vis. Res. 17 (8), 893–902.

Wechsler, D., 1981. Manual for the Adult Intelligence Scale-Revised.Psychological Corporation, New York.

Woods, S.W., 2003. Chlorpromazine equivalent doses for the neweratypical antipsychotics. J. Clin. Psychiatry 64, 663–667.