Measurement of Phase Gradients in the EEG for First Episode Psychosis

Proposal details

Title: Measurement of Phase Gradients in the EEG for First Episode Psychosis
Research Area(s): Schizophrenia and Allied Psychoses
Background: EEG measures of power and phase synchrony are related to symptom factors in First Episode Psychosis (FEP) subjects. We utilize a new dynamical measure that detects event-related episodes of long wavelength activity in the EEG. Long wavelength signals can be assessed by measuring phase-gradients in the EEG (Alexander et al., 2006). Long wavelength EEG signals have been shown to be related to a number of disorders: ADHD, Subjective Memory Complaint. This paper will explore the differences between FEP and matched controls on measures of long wavelength EEG.
Aims: To see if FEP subjects differ on measures of total long wavelength activity and specific measures of the direction of propagation of long wavelength activity. To relate these differences to clinical scores (PANNS).
Method: Data were acquired from 56 FEP subjects and 112 matched controls. PANSS were used to calculate 3-factor scores for Reality Distortion (RD), Psychomotor Poverty (PP) and Disorganization (Dis). EEG was measured during an Auditory Oddball task at 26 electrodes. The spatial patterns of phase gradients across the scalp were assessed using three, orthonormal basis functions: anterior-posterior, peripheral-central and right-left. Correlations of the relative phases with each of the basis functions are denoted rAP, rPC and rRL. Multiple t-tests were used to distinguish the two groups on the 4 measures (total wave activity, anterior-posterior waves, peripheral-central waves, right-left waves) for the frequencies 0.2 to 32Hz, from -200ms pre-stimulus to 800ms post-stimulus. The same measures were correlated with symptom factors to see which elements of long wavelength activity predict symptomology. To avoid type I errors, in each of these analyses, only 20 contiguous time/frequency points where p