Detection of EEG-Based Biomarkers of Chronic Low Back Pain

Overview

Chronic low back pain (CLBP) is a pervasive disorder affecting up to one-fifth of adults globally and is the single greatest cause of disability worldwide. Despite the high prevalence and detrimental impact of CLBP, its treatments and mechanisms remain largely unclear. Biomarkers that predict symptom progression in CLBP support precision-based treatments and ultimately aid in reducing suffering. Longitudinal brain-based resting-state neuroimaging of patients with CLBP has revealed neural networks that predict pain chronification and its symptom progression. Although early findings suggest that measurements of brain networks can lead to the development of prognostic biomarkers, the predictive ability of these models is strongest for short-term follow-up. Measurements of different neural systems may provide additional benefits with better predictive power. Emotional and cognitive dysfunction is common in CLBP, occurring at the behavioral and cerebral level, presenting a unique opportunity to detect prognostic brain-based biomarkers. Likewise, improvements in electroencephalogram (EEG) neuroimaging strategies have led to increased spatial resolution, enabling researchers to overcome the limitations of classically used neuroimaging modalities (e.g., magnetic resonance imaging \[MRI\] and functional MRI), such as high cost and limited accessibility. Using longitudinal EEG, this patient-oriented research project will provide a comprehensive neural picture of emotional, cognitive, and resting-state networks in patients with CLBP, which will aid in predicting symptom progression in CLBP. Through this award, the investigators will use modern EEG source analysis strategies to track biomarkers at baseline and 1- and 2-month follow-ups and their covariance with markers for pain and emotional and cognitive dysfunction. A 5-month follow up will also be used to only assess patient reported outcomes. In Aim 1, the investigators will identify and characterize differences in resting-state, emotional, and cognitive networks between patients with CLPB and age/sex-matched controls. In Aim 2, the investigators will identify within-subject changes across time and their relationship with clinical symptoms. In Aim 3, as an exploratory aim, the investigators will apply machine- and deep-learning strategies to detect a comprehensive signature of CLBP using EEG features from resting-state, emotional, and cognitive networks.