A ditor Hall cinations and Earl Vis al Processing Deficits in Schi ophrenia: Reference free ERPs D ring the Recognition of Words and FacesAuditory Hallucinations and Early Visual Processing Deficits in Schizophrenia: Reference-free ERPs During the Recognition of Words and FacesAuditory Hallucinations and Early Visual Processing Deficits in Schizophrenia: Reference free ERPs During the Recognition of Words and Facesy y g p g gJürgen Kayser 1,2 Craig E Tenke 1,2 Christopher J Kroppmann 1 Daniel M Alschuler 1 Shiva Fekri 1Jürgen Kayser 1,2, Craig E. Tenke 1,2, Christopher J. Kroppmann 1, Daniel M Alschuler 1, Shiva Fekri 1,g y , g , p pp , , ,

2 2 2 1 2Roberto Gil 2 Lars Jarskog 2 Jill M Harkavy-Friedman 2 Gerard E Bruder 1,250th Annual Meeting of the Roberto Gil , Lars Jarskog , Jill M. Harkavy-Friedman , Gerard E. Bruder ,50th Annual Meeting of theS i t f P h h i l i l R h (SPR)

g ySociety of Psychophysiological Research (SPR)

1 Di ision of Cogniti e Ne roscience Ne York State Ps chiatric Instit te Ne York NY;in Portland, Oregon, Sep. 29 – Oct. 3, 2010 1 Division of Cognitive Neuroscience, New York State Psychiatric Institute, New York, NY;g g , y , , ;2 D t t f P hi t C ll f Ph i i d S C l bi U i it N Y k NYPoster available in high resolution at 2 Department of Psychiatry, College of Physicians and Surgeons, Columbia University, New York, NYPoster available in high resolution at

http://psychophysiology cpmc columbia edu/mmedia/SPR2010/AH pdf http://psychophysiology.cpmc.columbia.edup y y, g y g , y, ,http://psychophysiology.cpmc.columbia.edu/mmedia/SPR2010/AH.pdf p p y p y gy p

Abstract Stimuli and Procedure Common Unknown Current Source DensitiesB k d E l d i l (ERP ) fl h hi h f i l b i i i

St u a d ocedu e CommonWords

UnknownFaces

Current Source DensitiesRecognition Memory (RM) Working Memory (WM)Background: Event-related potentials (ERPs) reflect the hierarchy of regional brain activity Words Faces

Continuous Recognition Memory TasksRecognition Memory (RM) Working Memory (WM)g p ( ) y g y

during the information processing sequence Reductions of early ERP components in CIRCLEContinuous Recognition Memory Tasks

A CB Dduring the information processing sequence. Reductions of early ERP components inschizophrenia have been inconsistently observed which may stem from differences in patients’

CIRCLE A CTEACHER TEACHERB Dschizophrenia have been inconsistently observed, which may stem from differences in patients TEACHERTEACHER TEACHER

symptoms. Importantly, N1 reductions to tones over left temporal cortex have been reported forEBEATy p p y p p

psychotic patients during periods of auditory hallucinations Methods: Taking advantage of two Exposure[ms]500 500

BEATControlspsychotic patients during periods of auditory hallucinations. Methods: Taking advantage of two

67 channel ERP data sets recorded during recognition memory (RM) and working memorynew item [ms]500 500REPLY Controls

Nonhallucinators67-channel ERP data sets recorded during recognition memory (RM) and working memory8 blocksTRACK Nonhallucinators

(WM) paradigms using words and faces, we compared right-handed patients who either lags: 8 items8 blocks

filler itemTRACK

Hallucinators( ) p g g , p g preported experiencing auditory hallucinations (n = 19/16 for RM/WM) or not (n = 32/35) and

lags: 8 items24 items F-W-W-F-W-F-F-Wfiller item

(never-repeated)VOTE

a uc ato sreported experiencing auditory hallucinations (n 19/16 for RM/WM) or not (n 32/35) andhealthy adults (n = 44/40) ERPs were transformed into reference free current source density

24 items or(never repeated)

SEARCHhealthy adults (n = 44/40). ERPs were transformed into reference-free current source density W-F-F-W-F-W-W-FSEARCH

P7 P8 P7 P8(CSD) waveforms (spherical spline surface Laplacian), which included a distinct N1 sink old item… P7 P8 P7 P8( ) ( p p p ),(approximate peak latency 150 ms) at inferior lateral-parietal sites (P7/8 P9/10) which was

old item…

TEACHER(approximate peak latency 150 ms) at inferior lateral parietal sites (P7/8, P9/10), which wasright lateralized for faces (N170) but strongly left lateralized for words CSDs were submitted to 912 trials total

TEACHERright-lateralized for faces (N170) but strongly left-lateralized for words. CSDs were submitted to 912 trials total

… 0 100 200 300 400 500 600 0 100 200 300 400 500 600 0 100 200 300 400 500 600 0 100 200 300 400 500 600unrestricted Varimax-PCA separately for each paradigm to quantify N1 sink activity. Results &

SEARCH

0 100 200 300 400 500 600 0 100 200 300 400 500 600 0 100 200 300 400 500 600 0 100 200 300 400 500 600p y p g q y y

Conclusions: In both paradigms N1 sink was substantially reduced in hallucinators compared /SEARCHConclusions: In both paradigms, N1 sink was substantially reduced in hallucinators compared

ith t l d h ll i t h did t diff f h th Th N1 i k • eight 114-trial blocks320 items/task

with controls and non-hallucinators, who did not differ from each other. These N1 sink • eight 114-trial blocks• each trial:reductions were most robust for words over the left hemisphere, suggesting an interference of • each trial:

gray stimulus frame (= fixation) -N1 i k N1 i kp , gg gactivity in temporoparietal language regions of hallucinators during early visual processing of

- gray stimulus frame (= fixation)b k f d (500 )

forced choice 184 filler N1 sink N1 sinkactivity in temporoparietal language regions of hallucinators during early visual processing ofd hi h t ib t t th i b l d fi it d i RM d WM t k

- yearbook face or common word (500 ms)f d h i ( / ld b tt 2 500 ISI)

button press184 filler

136 new/old pairs 0.1 µV/cm²words, which may contribute to their verbal memory deficits during RM and WM tasks. - forced choice (new/old button press, 2,500 ms ISI)p 136 new/old pairs 0.1 µV/cm

• response hand assignment new old +I t d ti

counterbalanced within subjectsnew old +

IntroductionSerial Position Tasks (Working Memory)• Reductions of early visual (e.g., Neuhaus et al 2010; Yeap et al 2006) and auditory (e.g., Salisbury et al 2010) ERP Serial Position Tasks (Working Memory)y ( g , ; p ) y ( g , y )

components in schizophrenia are often considered as evidence of sensory processing TEACHER REPLY VOTE REPLY P9 P10 P9 P10components in schizophrenia are often considered as evidence of sensory processing d fi it hi h t d h t f hi h i

TEACHER REPLY VOTE REPLY P9 P10 P9 P10deficits, which may represent an endophenotype for schizophrenia. forced choice

• Impairments of memory functions are one of the mostforced choicebutton press 0 100 200 300 400 500 600

Bruder et al, in pressImpairments of memory functions are one of the most significant cognitive deficits in schizophrenia and yet few

button press 0 100 200 300 400 500 600 0 100 200 300 400 500 600 0 100 200 300 400 500 600 0 100 200 300 400 500 600significant cognitive deficits in schizophrenia, and yet few t di i thi h dd d th iti l bl f 1 2 3 Latency [ms]studies in this area have addressed the critical problem of 1 2 3 Latency [ms]

biological and clinical heterogeneity in schizophrenia. encode 1 encode 2 encode 3 probeg g y p• Schizophrenic patients with auditory hallucinationsSchizophrenic patients with auditory hallucinations

have a greater verbal working memory (WM) deficit on 0 1500 3000 4500 6000 time [ms]have a greater verbal working memory (WM) deficit on 0 1500 3000 4500 6000 time [ms]the auditory word serial position task (WSPT) when Fig. 2. Grand mean reference-free CSD waveforms for controls, nonhallucinators, and hallucinators during RM and WM paradigms using words (A, C) or faces (B, D). Shown are inferior lateral-parietaly p ( )compared to non-hallucinators (Bruder et al in press) Recognition Memory (RM) Working Memory (WM)

Fig. 2. Grand mean reference free CSD waveforms for controls, nonhallucinators, and hallucinators during RM and WM paradigms using words (A, C) or faces (B, D). Shown are inferior lateral parietalsites where N1 sink (approximate peak latency 150 ms) was most prominent: over the left hemisphere for words (P7 P9) and over the right hemisphere for faces (P8 P10) Whereas virtually no differencecompared to non hallucinators (Bruder et al, in press). Recognition Memory (RM) Working Memory (WM) sites where N1 sink (approximate peak latency 150 ms) was most prominent: over the left hemisphere for words (P7, P9) and over the right hemisphere for faces (P8, P10). Whereas virtually no differencein N1 sink amplitude was observed between healthy controls and schizophrenic patients not reporting auditory hallucinations visual N1 sink amplitude appeared to be markedly reduced in hallucinatorsin N1 sink amplitude was observed between healthy controls and schizophrenic patients not reporting auditory hallucinations, visual N1 sink amplitude appeared to be markedly reduced in hallucinators.• Using two ERP data sets recorded from a large sample of schizophrenia patients and healthy

Hallucinators (n=19) Hallucinators (n=16)Using two ERP data sets recorded from a large sample of schizophrenia patients and healthy controls during recognition memory (RM) and working memory (WM) paradigms using words a 5 Nonhallucinators (n=32)

C l ( 44)100% Nonhallucinators (n=35)

C t l ( 40)** * CSD PCA Factor Loadings and Scores

controls during recognition memory (RM) and working memory (WM) paradigms using words d f b t d b ti ti t h t d i i dit at

a Controls (n=44)%

Controls (n=40)** CSD-PCA Factor Loadings and Scores]sm[ycnetaL ]sm[ycnetaLand faces, subgroups were created by separating patients who reported experiencing auditory

Da

90%*** ***g]sm[ycnetaL ]sm[ycnetaL

hallucinations in the past week (SAPS auditory hallucination item ≥ 1) from those who did not. D 4 t

90%*** *hallucinations in the past week (SAPS auditory hallucination item ≥ 1) from those who did not.

• The visual N1 to words is strongly left lateralized in healthy adults with a maximum over

al

dL] 4

rect

80% N1 sink0 50 100 150 200 250 300 350 400 0 50 100 150 200 250 300 350 400• The visual N1 to words is strongly left-lateralized in healthy adults, with a maximum over

th i f i l t l i t l i (i d i l t ) ra y[d

orr 80%

***N1 sink0 50 100 150 200 250 300 350 400 0 50 100 150 200 250 300 350 400the inferior lateral-parietal region (i.e., secondary visual cortex; e.g., Kayser et al 1999, 2007, 2009, 2010).

or ivity

3 tC ****** - -Th d d f f t ti l di f l ti ( li k d t id vio

siti 3

cent 70% TEACHER TEACHER

• The dependency of surface potentials on a recording reference location (e.g., nose, linked mastoids,

av Sens

erc

0.05 µV/cm² 148 0.05 µV/cm²TEACHER TEACHERaverage) and the definition and measurement of appropriate ERP components (e.g., specific time

ha S

2

Pe 60% 1410.05 µV/cm

+148 0.05 µV/cm

+

g ) pp p p ( g pwindows for peak or integral amplitudes) are two recurring problems in ERP research, which crucially eh 2 141 + +windows for peak or integral amplitudes) are two recurring problems in ERP research, which crucially affect component interpretation (e g polarity topography generator) and statistical analysis (e g B

e

Word Faceaffect component interpretation (e.g., polarity, topography, generator) and statistical analysis (e.g., Kayser & Tenke 2003; Tenke & Kayser 2005) B

0Word Face

0%Kayser & Tenke 2003; Tenke & Kayser 2005).Th li it ti b b bi i f f t d it (CSD) Word Face 1 2 3 1 2 3• These limitations can be overcome by combining reference-free current source density (CSD)transformations and temporal principal components analysis (PCA) to identify relevant, data-driven

Fig. 1. Mean (±SEM) performance accuracy during RM and WM tasks using words or faces. Overall, Controlsp p p p y ( ) y ,

components (Kayser &Tenke 2006a,b). g ( ) p y g g ,remembering faces was more difficult than remembering words. Patients performed more poorly than Controlscomponents (Kayser &Tenke 2006a,b).

Objecti eremembering faces was more difficult than remembering words. Patients performed more poorly thancontrols but hallucinators did not differ from nonhallucinators except for the WM word serial position taskObjective: controls, but hallucinators did not differ from nonhallucinators, except for the WM word serial position task.

• use CSD-PCA approach to quantify neuronal generator patterns underlying visual ERPs pp q y g p y g(i e N1 sink activity over lateral inferior-parietal regions) ERP Recording and Data Analysis Nose(i.e., N1 sink activity over lateral inferior-parietal regions)

N1 i k lit d f hi h i ti t ith d ith t ditERP Recording and Data Analysis Nose

2 6 1 8• compare N1 sink amplitude for schizophrenia patients with and without auditory Continuous DC (24 bit A/D 256 Hz) EEGs using an electrode cap 2 3 1 5-2.6 -1.8hallucinations to healthy controls • Continuous DC (24-bit A/D, 256 Hz) EEGs using an electrode cap,6 l i i di f (Bi S i) Left Right -2.3 -1.5y67 scalp sites, active recording reference (BioSemi) Left

MastoidRight

Mastoid• Amplifier drift eliminated by 2nd degree Polynomial high pass filter Non0 0 0 0

p y g y g p• Bipolar horizontal and vertical EOG; spatial SVD blink reduction;Participants Non- 0 0 0 00.0 0.0Bipolar horizontal and vertical EOG; spatial SVD blink reduction;

horizontal eye artifact reduction (epoched EEG linear regression)Participants hallucinators0.0 0.0

Means and Standard Deviations (SD) for Demographic and Clinical Variables horizontal eye artifact reduction (epoched EEG, linear regression)1 500 h 100 ti l b li ERP

p hallucinatorsMeans and Standard Deviations (SD) for Demographic and Clinical Variables

• 1,500 ms epochs, 100 ms pre-stimulus baseline, ERP averagesHallucinators Nonhallucinators Healthy Controls

+2 6 +1 8(artifact-free trials, correct responses only) low pass (12.5 Hz, -24 dB/oct)DSM-IV Criteria +2 3 +1 5Recognition Memory n = 19, 9 female n = 32, 14 female n = 44, 26 female +2.6 +1.8• reference-free current source densities (CSD; spherical splines surfaceSchizophrenia+2.3 +1.5

Working Memory n = 16, 8 female n = 35, 15 female n = 40, 22 female ( ; p pLaplacian; Perrin et al 1989) to sharpen ERP topographies and to eliminate

Schizophrenia g y , , ,

V i bl M SD M SD M SD Laplacian; Perrin et al., 1989) to sharpen ERP topographies and to eliminatevolume conducted activity from distant regions

Paranoid 18 18 Variable Mean SD Mean SD Mean SDvolume-conducted activity from distant regionsCSD t ll dj t d f N1 k l t f h t k i it P7 d P9 ( d) P7/8Undifferentiated 14 16 A ( ) 29.8 10.2 28.6 8.0 26.2 6.2

• CSDs temporally adjusted for N1 peak latency for each task, using sites P7 and P9 (word) or P7/8 Undifferentiated 14 16

1ldiRAge (years)

32.1 11.2 28.9 8.2 26.7 6.9and P9/10 (face), to create optimized epochs (-50 .. 400 ms) focused on N1 sink activity1laudiseR

32.1 11.2 28.9 8.2 26.7 6.914 2 2 6 14 3 2 4 16 3 2 1 ( ) p p ( ) y

• for each paradigm (RM, WM), CSD epochs focused on N1 (-50 .. 400 ms = 116 samples) submitted toSchizoaffective HallucinatorsEducation (years) 14.2 2.6 14.3 2.4 16.3 2.1

14 0 2 5 14 0 2 7 16 3 2 2 for each paradigm (RM, WM), CSD epochs focused on N1 ( 50 .. 400 ms 116 samples) submitted to unrestricted temporal principal components analysis (PCA) derived from the covariance matrixD d 8 5

Hallucinators14.0 2.5 14.0 2.7 16.3 2.272 8 27 7 75 0 20 2 75 3 20 0 unrestricted temporal principal components analysis (PCA) derived from the covariance matrix,

followed by Varimax rotation of covariance loadings (Kayser & Tenke 2003 2006a 2006b) to quantifyDepressed 8 5 Handedness (EHI-LQ) 72.8 27.7 75.0 20.2 75.3 20.0

77 9 25 0 78 3 18 3 74 7 19 7 followed by Varimax rotation of covariance loadings (Kayser & Tenke, 2003, 2006a, 2006b), to quantify l t tt d l i N1

Bipolar 7 8( )

77.9 25.0 78.3 18.3 74.7 19.7102 16 0 102 9 14 9 neuronal generator patterns underlying N1

hi f Psychosis NOS 4 3 Verbal IQ (WAIS) 102.7 16.0 102.9 14.9• CSD data: identify PCA components corresponding to visual N1 sink activity and submit factor scores at • no history of any

h th l Psychosis NOS 4 3 Verbal IQ (WAIS)

103.3 17.8 99.2 16.7 y p p g yrepresentative inferior lateral-parietal recording sites to repeated measures ANOVA with Grouppsychopathology or

l di dOnset age (years) 22.4 6.6 20.8 5.1 representative inferior lateral parietal recording sites to repeated measures ANOVA with Group(hallucinators nonhallucinators controls) Task (word face) Condition (RM: new old filler; WM:

neurology disorderMedication Status Onset age (years)24.2 6.3 20.4 5.4 (hallucinators, nonhallucinators, controls), Task (word, face), Condition (RM: new, old, filler; WM:

encode probe) Hemisphere (LH RH) and Site (P7/8 P9/10) as between and within subjects factors• matched to patients Unmedicated > 14 days 20 22 Fig. 3. Unrestricted PCA solutions using CSD subepochs optimized for N1 sink extraction. For each paradigm, one PCA factor uniquely corresponded to prominent N1 sinks over inferior lateral-parietal7.5 7.6 7.8 9.0 encode, probe), Hemisphere (LH, RH) and Site (P7/8, P9/10) as between- and within-subjects factorsB h i l d t f (RM iti it [d ] WM t t h

• matched to patients for core

Unmedicated > 14 days 20 22

Cl i 2 3sites (RM: 141 ms peak latency, 16.4% explained variance; WM: 148 ms, 14.4%). Across paradigms, N1 sinks were strongly left-lateralized for words (columns 1 and 3), and slightly right-lateralized forIllness duration (years) 7.5 7.6 7.8 9.0

7 9 9 0 8 5 8 9 • Behavioral data: performance accuracy measures (RM: sensitivity [dL]; WM: percent correct, chance for core demographics

Clozapine 2 3( p y, p ; , ) p g , g y ( ), g y g

faces (columns 2 and 4) In contrast to controls (top row) and nonhallucinators (middle row) hallucinators had markedly reduced N1 sinks with attenuated hemispheric asymmetries7.9 9.0 8.5 8.9

45 7 15 8 31 0 10 4 levels linearly scaled to 50% correct) submitted to repeated measures ANOVA with Group (hallucinators, demographics

Risperidone 6 7faces (columns 2 and 4). In contrast to controls (top row) and nonhallucinators (middle row), hallucinators had markedly reduced N1 sinks with attenuated hemispheric asymmetries.

Total BPRS a 45.7 15.8 31.0 10.443 4 13 2 33 3 10 6 y ) p p ( ,

nonhallucinators controls) and Task (word face) Condition (RM: lag [short long]; WM: serial positionp

Ol i 3 343.4 13.2 33.3 10.6

PANSS l a 37 4 13 8 26 9 8 7 nonhallucinators, controls), and Task (word, face), Condition (RM: lag [short, long]; WM: serial position [1 2 3]) as between and within subjects factors

Olanzapine 3 3ANOVA Means

PANSS general a 37.4 13.8 26.9 8.7[1, 2, 3]) as between- and within-subjects factors

Quetiapine 4 3 ANOVA Means141 14835.3 9.3 29.1 8.8

Ziprasidone 8 7141 148

PANSS positive a 19.2 7.4 12.2 5.4

Summary and ConclusionsZiprasidone 8 7 PANSS positive18.8 7.9 13.7 5.3 Summary and ConclusionsAripriprazole 8 6 Hallucinators (n=19) Hallucinators (n=16)

PANSS ti a 16.6 6.4 12.2 5.2• Patients and controls revealed highly comparable N1 sink peaks and topographies (Figs 2-3). -2.4

( )Nonhallucinators (n=32) -2.4

( )Nonhallucinators (n=35)

PANSS negative a

15.9 5.5 12.4 5.3 Patients and controls revealed highly comparable N1 sink peaks and topographies (Figs. 2 3).i 1 0 k l

Nonhallucinators (n=32)C t l ( 44)

Nonhallucinators (n=35)C t l ( 40)

15.9 5.5 12.4 5.3Note a Hallucinators differ significantly from Nonhallucinators(p # 02) • approximate 150 ms peak latency Word: LH >> RH Controls (n=44) Controls (n=40)Note. Hallucinators differ significantly from Nonhallucinators(p # .02).

R f

pp p y• inferior lateral-parietal maximum (P7/8, P9/10) Word: LH >> RH -2.0 ** -2.0

References Supported by grant MH066597 from the National Institute of Mental Health (NIMH)inferior lateral parietal maximum (P7/8, P9/10)

• concisely summarized by temporal PCA Face: LH < RH e * e

Bruder GE Alschuler DM Kroppmann CJ Fekri S Gil R Jarskog LF Harkavy-Friedman Kayser J Tenke CE Gil RB Bruder GE (2009) Stimulus- and response-locked neuronal• concisely summarized by temporal PCA Face: LH RH

ore

ore

Bruder GE, Alschuler DM, Kroppmann CJ, Fekri S, Gil R, Jarskog LF, Harkavy FriedmanJM, Goetz R, Kayser J, Wexler BE (in press). Heterogeneity of Auditory Verbal Working

Kayser J, Tenke CE, Gil RB, Bruder GE (2009). Stimulus and response locked neuronalgenerator patterns of auditory and visual word recognition memory in schizophrenia. co -1.6 co -1.6 **• Across memory paradigms, hallucinators had markedly reduced N1 sinks for words and faces , , y , ( p ) g y y g

Memory in Schizophrenia. J. Abnorm. Psychol.g p y g y pInt. J. Psychophysiol. 73(3):186-206. Sc 1.6 Sc 1.6

*Across memory paradigms, hallucinators had markedly reduced N1 sinks for words and faces compared to healthy controls and nonhallucinators particularly over the left hemisphere (Fig 4)Ford JM, Roach BJ, Jorgensen KW, Turner JA, Brown GG, Notestine R, et al (2009).

Tuning in to the voices: a multisite FMRI study of auditory hallucinations SchizophrKayser J, Tenke CE, Kroppmann CJ, Fekri S, Alschuler DM, Gates NA, Gil R, Harkavy-

Friedman JM Jarskog LF Bruder GE (2010) Current source density (CSD) old/new or

* ( ) or compared to healthy controls and nonhallucinators, particularly over the left hemisphere (Fig. 4).

Tuning in to the voices: a multisite FMRI study of auditory hallucinations. Schizophr.Bull 35(1):58-66

Friedman JM, Jarskog LF, Bruder GE (2010). Current source density (CSD) old/neweffects during recognition memory for words and faces in schizophrenia and in healthy ct

o

-1.2 * *( ) cto

-1.2• In contrast both hallucinators and nonhallucinators performed more poorly than controls across Bull. 35(1):58 66.Hubl D, Koenig T, Strik WK, Garcia LM, Dierks T (2007). Competition for neuronal

effects during recognition memory for words and faces in schizophrenia and in healthyadults. Int. J. Psychophysiol. 75(2):194-210. Fa

c 1.2** * Fa

c 1.2

*( ) *( ) *• In contrast, both hallucinators and nonhallucinators performed more poorly than controls across , g , , , ( ) p

resources: how hallucinations make themselves heard. Br. J. Psychiatry 190:57-62.y p y ( )

Neuhaus AH, Karl C, Hahn E, Trempler NR, Opgen Rhein C, Urbanek C, Hahn C, Ta TM, F * F *( ) *( ) *( ) *paradigms and both word and face tasks; however, hallucinators performed more poorly than Kayser J, Bruder GE, Friedman D, Tenke CE, Amador XF, Clark SC, Malaspina D, Gorman

JM (1999) Brain e ent related potentials (ERPs) in schi ophrenia d ring a ordDettling M (2010). Dissection of early bottom-up and top-down deficits during visualattention in schizophrenia Clin Neurophysiol [Epub ahead of print] -0 8 -0 8 *

p g ; , p p ynonhallucinators during the word working memory paradigm (Fig 1) JM (1999). Brain event-related potentials (ERPs) in schizophrenia during a word

recognition memory task Int J Psychophysiol 34(3):249-265attention in schizophrenia. Clin. Neurophysiol. [Epub ahead of print]

Perrin F Pernier J Bertrand O Echallier JF (1989) Spherical splines for scalp potential0.8 0.8nonhallucinators during the word working memory paradigm (Fig. 1).

recognition memory task. Int. J. Psychophysiol. 34(3):249-265.Kayser J, Tenke CE (2003). Optimizing PCA methodology for ERP component

Perrin F, Pernier J, Bertrand O, Echallier JF (1989). Spherical splines for scalp potentialand current density mapping. Electroencephalogr. Clin. Neuro-physiol. 72(2):184-187. • These findings are consistent with N1 reductions to tones in psychotic patients during periods of y , ( ) p g gy p

identification and measurement: theoretical rationale and empirical evaluation. Clin.y pp g p g p y ( )

[1990 Corrigenda EEG 02274. Electroencephalogr. Clin. Neurophysiol. 76:565] Word Face-0 4 Word Face-0 4• These findings are consistent with N1 reductions to tones in psychotic patients during periods of

Neurophysiol. 114(12):2307-2325.K J T k CE (2006 ) P i i l t l i f L l i f

Tenke CE, Kayser J (2005). Reference-free quantification of EEG spectra: combiningt d it (CSD) d f i i l t l i (fPCA)

Word Face0.4 Word Face0.4auditory hallucinations as opposed to symptom-free intervals (Hubl et al 2007) and fMRI evidence Kayser J, Tenke CE (2006a). Principal components analysis of Laplacian waveforms as ageneric method for identifying ERP generator patterns: I Evaluation with auditory

current source density (CSD) and frequency principal components analysis (fPCA).Clin Neurophysiol 116(12):2826-2846

y pp y p ( )revealing differences between hallucinators and nonhallucinators (Ford et al 2007; Wible et al 2009)generic method for identifying ERP generator patterns: I. Evaluation with auditory

oddball tasks. Clin. Neurophysiol. 117(2):348-368.Clin. Neurophysiol. 116(12):2826-2846.

Wible CG, Lee K, Molina I, Hashimoto R, Preus AP, Roach BJ, et al (2009). fMRI activity LH RH LH RH LH RH LH RHrevealing differences between hallucinators and nonhallucinators (Ford et al 2007; Wible et al 2009).oddba tas s C eu op ys o ( ) 3 8 368Kayser J, Tenke CE (2006b). Principal components analysis of Laplacian waveforms as a

b e CG, ee , o a , as oto , eus , oac J, et a ( 009) act tycorrelated with auditory hallucinations during performance of a working memory task: • Results provide further neurophysiological evidence of early visual processing deficits in schizophrenia generic method for identifying ERP generator patterns: II. Adequacy of low-density

ti t Cli N h i l 117(2) 369 380data from the FBIRN consortium study. Schizophr. Bull. 35(1):47-57.

Y S K ll SP S h t P M E J itt DC G H Th k JH F JJ Fig 4 Pooled means (±SEM) of N1 sink (mean scores of factors 141 [RM] and 148 [WM]) at inferior lateral-parietal sites (P7/8 P9/10) with ANOVA significance levels of simple group comparisons for• Results provide further neurophysiological evidence of early visual processing deficits in schizophrenia.

estimates. Clin. Neurophysiol. 117(2):369-380.Kayser J Tenke CE Gates NA Bruder GE (2007) Reference-independent ERP old/new

Yeap S, Kelly SP, Sehatpour P, Magno E, Javitt DC, Garavan H, Thakore JH, Foxe JJ(2006) Early visual sensory deficits as endophenotypes for schizophrenia: high-

Fig. 4. Pooled means (±SEM) of N1 sink (mean scores of factors 141 [RM] and 148 [WM]) at inferior lateral parietal sites (P7/8, P9/10) with ANOVA significance levels of simple group comparisons foreach task and hemisphere For both paradigms overall main effects of Group (RM: F = 3 20 p = 04; WM: F = 2 95 p = 06) originated from reduced N1 sinks in hallucinators compared to• While these findings support the value of a symptom-based approach for dealing with the heterogeneity of Kayser J, Tenke CE, Gates NA, Bruder GE (2007). Reference-independent ERP old/new

effects of auditory and visual word recognition memory: joint extraction of stimulus-(2006). Early visual sensory deficits as endophenotypes for schizophrenia: high-density electrical mapping in clinically unaffected first-degree relatives. Arch. Gen.

each task and hemisphere. For both paradigms, overall main effects of Group (RM: F[1,92] = 3.20, p = .04; WM: F[1,88] = 2.95, p = .06) originated from reduced N1 sinks in hallucinators compared tot l (RM 02 WM 02) d h ll i t (RM 06 WM 04) h t l d h ll i t did t diff i N1 i k lit d

• While these findings support the value of a symptom-based approach for dealing with the heterogeneity of hi h i i l N1 d ti t t bi l i l k f hi h i effects of auditory and visual word recognition memory: joint extraction of stimulus

and response-locked neuronal generator patterns. Psychophysiology 44(6):949-967.density electrical mapping in clinically unaffected first degree relatives. Arch. Gen.Psychiatry 63(11):1180-1188. controls (RM: p = .02; WM: p = .02) and nonhallucinators (RM: p = .06; WM: p = .04), whereas controls and nonhallucinators did not differ in N1 sink amplitude.schizophrenia, visual N1 reductions may not represent a biological marker for schizophrenia per se.