Critical Evaluation of Auditory Event-Related Potential Deficits in Schizophrenia: Evidence From Large-Scale Single-Subject Pattern Classification

Abstract

Event-related potential (ERP) deficits associated with auditory oddball and click-conditioning paradigms are among the most consistent findings in schizophrenia and are discussed as potential biomarkers. However, it is unclear to what extend these ERP deficits distinguish between schizophrenia patients and healthy controls on a single-subject level, which is of high importance for potential translation to clinical routine. Here, we investigated 144 schizophrenia patients and 144 matched controls with an auditory click-conditioning/oddball paradigm. P50 and N1 gating ratios as well as target-locked N1 and P3 components were submitted to conventional general linear models and to explorative machine learning algorithms. Repeatedmeasures ANOVAs revealed significant between-group differences for the oddball-locked N1 and P3 components but not for any gating measure. Machine learning-assisted analysis achieved 77.7% balanced classification accuracy using a combination of target-locked N1 and P3 amplitudes as classifiers. The superiority of machine learning over repeated-measures analysis for classifying schizophrenia patients was in the range of about 10% as quantified by receiver operating characteristics. For the first time, our study provides large-scale single-subject classification data on auditory click-conditioning and oddball paradigms in schizophrenia. Although our study exemplifies how automated inference may substantially improve classification accuracy, our data also show that the investigated ERP measures show comparably poor discriminatory properties in single subjects, thus illustrating the need to establish either new analytical approaches for these paradigms or other paradigms to investigate the disorder.

The development of the N1 and N2 components in auditory oddball paradigms: a systematic review with narrative analysis and suggested normative values

“…The most standard stimuli used was a pure-tone of 1000 Hz at an intensity level range from 55 to 109 dB above hearing threshold, with at least 200 Hz of difference to the deviant stimuli, except in Demiralp et al (1999), Wang et al (2005) and Č eponien_ e et al (2008) studies. Others used duration auditory oddball (Shelley et al 1999;Segalowitz et al 2001;Bortoletto et al 2011;Wetzel et al 2011;Neuhaus et al 2013) and few intensity auditory oddball (Anderer et al 1996(Anderer et al , 1998aWang and Wang 2001;Wang et al 2005;Barry et al 2006). Only two papers elicited N1 and N2 with speech stimuli (Henkin et al 2002;Toscano et al 2010), due to the scarce data and type of stimuli they were not considered in the regression analysis.…”

“…This fact, suggests that in the last decades N1 and N2 components have been studied as possible endophenotypes or bio-markers for different disorders and in some cases for recovery index: namely, in schizophrenia (Hegerl et al 1988;Haig et al 1997;Potts et al 1998a;Shelley et al 1999;Ford et al 2001;Brown et al 2002;Gilmore et al 2005;Neuhaus et al 2013), dementia (Verma et al 1989), Alzheimer (Sumi et al 2000), epilepsy and AED effects (Akaho 1996;Ford et al 2001;Lucchesi et al 2003), alcohol (Brigham et al 1997;Cohen et al 2002) and substance abuse (Tarter et al 1995;Brigham et al 1997), psychosis (Valkonen-Korhonen et al 2003), panic disorder (Wise et al 2009), Parkinson's disease (Wright et al (1996), ADHD (Johnstone et al 2001;Tsai et al 2012), stroke (Hirata et al 1996) and vascular cognitive impairment (van Harten et al 2006), myotonic dystrophy (Kazis et al 1996;Tanaka et al 2012), head injury (Reinvang et al 2000;Segalowitz et al 2001), (Guney et al 2009), amyotrophic lateral sclerosis (Ogawa et al 2009) and multiple sclerosis (Whelan et al 2010), more recently fragile X syndrome treatment (Schneider et al 2013), and bipolar disorder (Hamm et al 2013). …”

Early auditory processing evoked potentials (N100) show a continuum of blunting from clinical high risk to psychosis in a pediatric sample

“…Thus, pathology affecting these layers and projections may reduce N100 responses. Notably, in primarily adult studies, N100 amplitude is reduced in first episode and chronic schizophrenia (SZ) patients, both medicated and unmedicated, and in their relatives (Brockhaus-Dumke et al, 2008; Brown et al, 2002; Hsieh et al, 2012; Mukundan, 1986; Neuhaus et al, 2013; Rosburg et al, 2008; Turetsky et al, 2008). Studies have differed as to whether the N100 is decreased in CHR and whether it decreases further with progression to psychosis (Brockhaus-Dumke et al, 2008; Hsieh et al, 2012; van Tricht et al, 2015).…”

Auditory prediction errors as individual biomarkers of schizophrenia

“…Moreover, it departs methodologically from different studies in that it takes the whole spatiotemporal information in the EEG data as potential features for the classifiers. Neuhaus et al (2014) used a combined click-conditioning and auditory oddball paradigm with square wave standards and sinusoidal deviants, explicitly targeting the P50, N100 and P300 components. Our approach yielded a greater balanced classification accuracy (up to 80.48% in our study compared to 77.7%), and our ROC analysis indicated that the left gap paradigm provides greater discrimination between groups with an AUC of 0.87 outperforming 0.737 reported in (Neuhaus et al, 2014).…”