Validation of a suite of ERP and QEEG biomarkers in a pre-competitive, industry-led study in subjects with schizophrenia and healthy volunteers.

OBJECTIVE: Complexity and lack of standardization have mostly limited the use of event-related potentials (ERPs) and quantitative EEG (QEEG) biomarkers in drug development to small early phase trials. We present results from a clinical study on healthy volunteers (HV) and patients with schizophrenia (SZ) that assessed test-retest, group differences, variance, and correlation with functional assessments for ERP and QEEG measures collected at clinical and commercial trial sites with standardized instrumentation and methods, and analyzed through an automated data analysis pipeline. METHODS: 81 HV and 80 SZ were tested at one of four study sites. Subjects were administered two ERP/EEG testing sessions on separate visits. Sessions included a mismatch negativity paradigm, a 40 Hz auditory steady-state response paradigm, an eyes-closed resting state EEG, and an active auditory oddball paradigm. SZ subjects were also tested on the Brief Assessment of Cognition (BAC), Positive and Negative Syndrome Scale (PANSS), and Virtual Reality Functional Capacity Assessment Tool (VRFCAT). RESULTS: Standardized ERP/EEG instrumentation and methods ensured few test failures. The automated data analysis pipeline allowed for near real-time analysis with no human intervention. Test-retest reliability was fair-to-excellent for most of the outcome measures. SZ subjects showed significant deficits in ERP and QEEG measures consistent with published academic literature. A subset of ERP and QEEG measures correlated with functional assessments administered to the SZ subjects. CONCLUSIONS: With standardized instrumentation and methods, complex ERP/EEG testing sessions can be reliably performed at clinical and commercial trial sites to produce high-quality data in near real-time.

Multimodal Prediction of Alzheimer's Disease Severity Level Based on Resting-State EEG and Structural MRI.

While several biomarkers have been developed for the detection of Alzheimer's disease (AD), not many are available for the prediction of disease severity, particularly for patients in the mild stages of AD. In this paper, we explore the multimodal prediction of Mini-Mental State Examination (MMSE) scores using resting-state electroencephalography (EEG) and structural magnetic resonance imaging (MRI) scans. Analyses were carried out on a dataset comprised of EEG and MRI data collected from 89 patients diagnosed with minimal-mild AD. Three feature selection algorithms were assessed alongside four machine learning algorithms. Results showed that while MRI features alone outperformed EEG features, when both modalities were combined, improved results were achieved. The top-selected EEG features conveyed information about amplitude modulation rate-of-change, whereas top-MRI features comprised information about cortical area and white matter volume. Overall, a root mean square error between predicted MMSE values and true MMSE scores of 1.682 was achieved with a multimodal system and a random forest regression model.