Auditory P3a and P3b neural generators in schizophrenia: An adaptive sLORETA P300 localization approach

Abstract

The present study investigates the neural substrates underlying cognitive processing in schizophrenia (Sz) patients. To this end, an auditory 3-stimulus oddball paradigm was used to identify P3a and P3b components, elicited by rare-distractor and rare-target tones, respectively. Event-related potentials (ERP) were recorded from 31 Sz patients and 38 healthy controls. The P3a and P3b brain-source generators were identified by time-averaging of low-resolution brain electromagnetic tomography (LORETA) current density images. In contrast with the commonly used fixed window of interest (WOI), we proposed to apply an adaptive WOI, which takes into account subjects' P300 latency variability. Our results showed different P3a and P3b source activation patterns in both groups. P3b sources included frontal, parietal and limbic lobes, whereas P3a response generators were localized over bilateral frontal and superior temporal regions. These areas have been related to the discrimination of auditory stimulus and to the inhibition (P3a) or the initiation (P3b) of motor response in a cognitive task. In addition, differences in source localization between Sz and control groups were observed. Sz patients showed lower P3b source activity in bilateral frontal structures and the cingulate. P3a generators were less widespread for Sz patients than for controls in right superior, medial and middle frontal gyrus. Our findings suggest that target and distractor processing involves distinct attentional subsystems, both being altered in Sz. Hence, the study of neuroelectric brain information can provide further insights to understand cognitive processes and underlying mechanisms in Sz.

Introduction

Alterations in cognitive processing in schizophrenia (Sz) have long been assessed using electroencephalographic (EEG) recordings (Roach and Mathalon, 2008). In particular, it is usual to obtain the event-related potentials (ERP) as the average of EEG epochs time-locked to repeated external stimulus or events. Reduced P300 amplitude during an auditory oddball paradigm is one of the most consistent findings in schizophrenia (Sz) (Bramon et al., 2004); however, the neural bases of this amplitude reduction are incompletely understood. In this regard, the analyses focused on the localization of neural generators can contribute to elucidate possible sources of altered information processing in Sz (Mulert et al., 2004).

The oddball paradigm is a common experimental design used in ERP analyses to obtain the P300 wave. The 3-stimulus variant of the auditory-oddball paradigm is characterized by infrequent-distractor stimuli interspersed randomly into a sequence of frequent-standard and rare-target. This paradigm allows the examination of cognitive processing as response to both relevant and irrelevant stimuli (Polich, 2007). The resulting P300 wave includes two components: the P3a, evoked by distractor stimuli for which no subject-response is expected; and the P3b, elicited by target stimuli for which the subject is instructed to respond. The neural processing as a response to auditory distractor tones has been related to bottom-up attentional mechanisms; hence, P3a may be generated whether sufficient attentional focus is engaged. In contrast, P3b seems to be related to conscious top-down target processing, likely contributing to processing the stimulus information and performing cognitive response (Polich, 2007, Strobel et al., 2008). In previous reports, we found a blunted ERP modulation in Sz as response to both target (Bachiller et al., 2014) and distractor (Bachiller et al., 2015b) tones during an oddball paradigm. Thus, the analysis of the differences in neural pattern generators between Sz patients and healthy controls becomes an interesting research topic to clarify the neural substrate of reduced P300 amplitude in Sz.

Source imaging techniques may help to detect neural generators that contribute to the scalp recorded ERPs, resulting in an acceptable compromise between spatial and temporal resolutions. The inverse solution (i.e. the computation of 3-D intracerebral images of electric neuronal activity based on scalp-recorded EEG) would provide useful information on the time course and localization of brain functions (Pascual-Marqui, 2002). There is no unique solution to the inverse problem; nevertheless, the low-resolution brain electromagnetic tomography (LORETA) is one of the most reliable methods for localizing ERP electrical activity and it is associated with relatively low error rates (Pascual-Marqui, 2002, Jung et al., 2012). LORETA has been widely used for source localization in psychiatric disorders, such as Sz or depression (Kawasaki et al., 2004, Mientus et al., 2002). Moreover, auditory P3a and P3b source localization has been previously addressed in healthy controls using LORETA and functional magnetic resonance imaging (fMRI). P3a generators are localized in anterior cingulate, frontal area and parietal cortices (Polich, 2007, Strobel et al., 2008, Volpe et al., 2007), whereas P3b sources include a more distributed network, involving superior and medial temporal, posterior parietal, hippocampal, cingulate and frontal structures (Polich, 2007, Strobel et al., 2008, Volpe et al., 2007, Wronka et al., 2012).

It is noteworthy that the previous LORETA findings are influenced by one important technical shortcoming: although ERP analyses show a considerable inter-subject variability of P300 latency (Campanella et al., 1999), LORETA source imaging studies commonly used a large fixed post-stimulus window of interest (WOI), like [250 500] ms (Higuchi et al., 2008, Kawasaki et al., 2007, Sumiyoshi et al., 2006), [280 450] ms (Wang et al., 2003, Wang et al., 2010), [240 420] ms (Pae et al., 2003), [227 383] ms (Volpe et al., 2007), or [400 700] ms (Wronka et al., 2012).

In this study we proposed a new LORETA approach based on P300 wave (‘P300 latency adaptive WOI’) to properly localize P300 brain-source generators in each subject. To the best of our knowledge, this is the first study that analyzes the sources of both P3a and P3b in Sz using LORETA. Hence, this research is aimed at: (i) analyzing the performance of the adaptive WOI method in comparison to conventional fixed WOI analysis; and (ii) applying the adaptive WOI method to analyze the differences of auditory P3a and P3b underlying cortical sources between Sz patients and healthy controls.