Pii: s0167-8760(02)00035-

International Journal of Psychophysiology 45 (2002) 95–178 as a result of treatment with citalopram. The results suggest that potentials are different in subjects with dyslexia in comparison to both the ‘slowed thinking’ and altered neuroanatomy may be normal readers in the above mentioned areas. The EEG at rest reversed by successful treatment of major depression.
shows an increase of theta activity on parietal regions and thecerebral evoked potentials show an increase in latency and a reduc- NEWBORN ERPS PREDICT LATER LANGUAGE SKILLS IN tion in amplitude. These differences are found in the unimodal and CHILDREN WITH AND WITHOUT FAMILIAL RISK FOR multimodal cerebral areas. These observations derive mainly from experimental conditions in which the normal subjects or with dys-lexia view passively letters, words or symbols. This study presentsthe brain electrical responses when a normal subject and with dys- lexia is actively engaged in reading aloud letters, either externally Departmert of Psychology, University of Jyvaskyla, Finland paced or self-paced. Forty two normal children and 18 with devel-opmental dyslexia ranging from 8 to 10 years old participated in Early identification of dyslexia would facilitate well directed reme- the study. The EEG was recorded from Fz, Cz, Pz, Oz, right and diation even before language problems are typically diagnosed.
left pre-central, P3, P4, T3, T4, referenced to linked mastoids.
Currently, few methods are available for identifying infants who EMGs from the right forearm and lips, ECG, PNH and voice will later have difficulties in their language development, and who response were also recorded. The reading performance of children thus would benefit from interventions. Our results show that new- with dyslexia were significantly lower (P-0.0001) than the nor- born event-related potentials (ERPs) are associated with later lan- mal children during both externally and self-paced reading. Those guage skills, suggesting that ERPs have potential to be developed dyslexic children with low reading quotient had lower reading per- as a diagnostic tool. The ERPs to synthetic consonant–vowel syl- formance during externally paced reading than during self-paced lables (ybay, yday, ygay) were measured from 26 newborns with reading. Student t-test and variance analysis showed that signifi- familial risk for dyslexia and 23 control infants participating in the cant differences were present between normal and dyslexic sub- ¨ Longitudinal Study of Dyslexia. The syllables were pre- jects and between reading aloud letters externally paced and sented with equal probability and with 3910–7285 ms interstimu- self-paced. The potentials most significantly affected were P2 lus intervals. Analyses of ERPs using the latencies identified with (270–330 ms), N3 (310–375 ms), N4 (380–650 ms) and the Ber- principal component analysis revealed clearest hemispheric differ- eitschafts-potential in Fz, Cz, Pz, RPC, LPC.
ences between groups in processing of speech stimuli at the latency540–630 ms: In the responses to ygay, the polarity shift from the AUTOMATIC MEASUREMENT OF ERPS IN NORMAL AND major positive peak toward the later negative deflection occurred DYSLEXIC CHILDREN BY MEANS OF A NON-LINEAR later in the at-risk group (clearest at the right hemisphere). Cor- relation and regression analyses showed that this pattern at the lefthemisphere was related to the poorer language skills in these same Silvia Casarottoa, Anna M. Bianchia, Sergio Ceruttia, Giuseppe A.
children at the ages between 2 y 6 m and 5 years. The similar pattern at the right hemisphere was associated with poorer lan- of Biomedical Engineering, Polytechnic University, guage and visuo-spatial skills at the later stages of development.
These results suggest that the differential role of the hemispheres of Child and Adolescent Neuropsychiatry, Az. Osp.
in auditory speech processing predicts later language skills, and G. Salvini, Rho Hospital, Rho, Milan, Italy could thus be used in early identification of children at risk forlanguage problems.
The development of methods for automatic detection of peaks andtroughs is a great challenge in order to cope with the intra- and inter-individual variability of ERPs and with that due to subjectiveevaluation of the experimenter.
Giuseppe A. Chiarenza, Giulia Frasson, Igor De Marchi The present study was performed on a population of normal and Department of Child and Adolescent Neuropsychiatry, Rho Hos- dyslexic children aged 8 and 9 years. The EEG was unipolarly recorded from Fz, Cz, Pz, Oz, C3, C4, P3, P4, T3, T4, referencedto linked mastoids; the EOG was bipolarly recorded. The method The psychophysiological approach to reading processes is typically was applied to four different paradigms, designed to investigate based on three types of studies: quantitative EEG, sensory evoked the reading processes in normal and dyslexic children. Two para- potentials and cognitive potentials. Considered all together, these digms, defined as passive, consisted in viewing letters or symbols studies show that during reading, numerous cerebral areas of both without linguistic content. The active paradigms consisted of read- hemispheres are activated. The pattern of activation depends on ing aloud letters after externally or self-paced action. A non-linear the type of stimuli, simple or complex, on modality presentation, alignment technique based on dynamic time warping was used: it visual or auditory, on experimental conditions, on the involvement aligns the samples of the signals on the basis of their morpholog- of the subject, active or passive, and on the clinical diagnosis. The ical similarity through a local compression and extension of their cerebral areas involved in the reading processes are the parieto- temporal axes. This similarity was measured from the amplitude temporo-occipital areas and the motor areas. The cerebral electrical and the first derivative of the signals. An ERP template was activity both spontaneous and evoked, the sensory and cognitive obtained for each group and for each task. The automatic meas- International Journal of Psychophysiology 45 (2002) 95–178 urement of the latencies and amplitudes of the relevant waves was percentage was 65.9% in normal and 72.5% in dyslexic children.
obtained through the alignment of the single subject’s ERP with With the application of PCA, the number of artefact-free trials the corresponding template. Templates obtained during active tasks increased of 155 and 179% in normal children during passive and were characterized by significantly greater amplitudes with respect active conditions, respectively. In dyslexic children the increase per to that recorded during passive tasks in most of the cerebral areas cent was of 150 and 189% during passive and active conditions, recorded. In particular, during self-paced task there was a reduction respectively. The achievements of the method are: a reduction of of latency of ERPs’ components before 200 ms. Dyslexic children the experiment time and a great improvement in the quality of had a significant increase of latencies of ERPs’ components after ERPs, without modifications of latencies and amplitudes.
200 ms, mainly in parietal and central areas.
This method applied to ERPs, extends the previous applications APPLICATIONS OF THE ‘MANGINA-TEST’ IN THE CLINI- of non-linear alignment techniques to short latency potentials.
CAL INVESTIGATION OF CHILDREN AND ADOLESCENTS Despite of the great interindividual variability that characterizes WITH NEUROPSYCHIATRIC PATHOLOGIES AS COM- ERPs in children, the method succeeds in reducing the temporal differences between the signals and in making easier their com-parison. This non-linear method overcomes other heuristic tech- niques that require great experience of the experimenter, are time Department of Child and Adolescent Neuropsychiatry, Rho Hos- consuming and fail in the automatic recognition of the relevant The ‘Mangina-Test’ is used for the diagnosis of varying degrees PRINCIPAL COMPONENT ANALYSIS FOR REDUCTION OF of ‘analytical-specific visual perception’ and was standardized in North America as a neuropsychometric diagnostic tool for specific learning abilities and disabilities (Mangina, 1981, 1994, 1998; Erl-baum Publishers, New Jersey, USA).
Silvia Casarottoa, Anna Bianchia, Sergio Ceruttia, Giuseppe A.
The purpose of our present research was to apply the ‘Mangina- Test’ to investigate the ‘analytical-specific perceptual skills’ of of Biomedical Engineering, Polytechnic University, pathological and normal children and adolescents in Italy. The con- trol group consisted of 58 children, 10 girls and 48 boys which of Child and Adolescent Neuropsychiatry, Az. Osp.
age ranged from 7.10 years to 14 years, mean age 10.6. The per- G. Salvini, Rho Hospital, Rho, Milan, Italy formance of our normal sample fell within the normal four sub-categories of the ‘Mangina-Test’: 74.1% above average perceptual The recording of ERPs in normal and pathological children is abilities, superior and very superior ‘analytical-specific perceptual always a difficult and time-consuming task, because they cannot abilities’ and 25.9% average abilities similar. The pathological sub- refrain from blinking or moving, due to developmental or patho- jects to whom the ‘Mangina-Test’ was applied consisted of 130 logical reasons. We present an application of principal component subjects, 89 boys and 41 girls, within an age range between 5.6 analysis (PCA) for removing ocular artefacts from single EEG and 16.0 years, (mean age 8.7) and divided in 8 diagnostic groups: recordings of ERPs in normal and dyslexic children.
generalized learning disabilities (Ns17), specific learning disa- The EEG was unipolarly recorded from Fz, Cz, Pz, Oz, C3, C4, bilities (Ns27), language disorders (Ns12), attention deficit dis- P3, P4, T3, T4, referenced to linked mastoids; the EOG was bipo- orders (Ns6), mixed developmental disorders (Ns8), not larly recorded. The method was applied to four different para- otherwise specified learning disabilities (Ns4), mental retardation digms, designed to investigate the reading processes in normal and (Ns17), motor disturbances (Ns3). ‘Analytical-specific percep- dyslexic children. Two paradigms, defined as passive, consisted in tual disabilities’ were present in 72.3% of all considered patholo- viewing letters of the alphabet or symbols without linguistic con- gies, ranging from severe to moderate perceptual disabilities. Out tent. The active paradigms consisted in reading aloud letters after of this percentage, 100% of children with mental retardation externally paced or self-paced action. The method described below revealed perceptual disabilities, followed by children with gener- was applied to every single trial: PCA was applied on a matrix alized learning disabilities (89.5%) and mixed developmental dis- containing EOG and all EEG recordings. Correlation coefficients orders (80.0%) and with equal percentage (75.0%) children with between EOG, the first (c1) and the second (c2) principal com- attention deficit disorders and motor developmental disorders. The ponent were computed. The method had effect on the recorded comparison of the diagnostic categories of the ‘Mangina-Test’ trials only in the following two conditions: the first principal com- between the pathological subjects and normal controls showed ponent was subtracted from the original trial if c100.9; the second highly significant differences (P-0.001) between the two groups: principal component was subtracted if c1-0.9 and c200.95. The generalized learning disabilities (zsy6.37), specific learning dis- number of trials with artefacts decreased with age: 8-year-old chil- abilities (zsy6.52), language disorders (zsy3.79), attention dren had 1.2 times the number of artefacted trials recorded in 9- deficit disorders (zsy4.12), mixed developmental disorders (zs year-old children. The active paradigms produced 4% more y5.11), not otherwise specified learning disabilities (zsy4.15), artefacted trials than passive paradigms. The percentage of arte- mental retardation (zsy6.56), motor disturbances (zsy3.41).
facted trials during passive paradigms was 57.1% in normal chil- The results of this investigation clearly testify as to the clinical dren and 56.2% in dyslexics; during active paradigms the usefulness of the ‘Mangina-Test’ and also confirm its ‘culture-free’

Source: http://www.giuseppechiarenza.it/2002/IOP02.pdf