Graph Models for Chronic Pain Detection through EEG Biomarkers
Can we Objectively Measure Chronic Pain?
Despite significant strides in healthcare and clinical technology, the methods for assessing pain, particularly chronic pain, have remained largely stagnant. The complexity of identifying specific neuro markers has hindered the progress, leaving clinical assessments heavily reliant on subjective measures and traditional rating scales.
Our preliminary analysis introduces a novel approach to chronic pain detection utilizing electroencephalographic (EEG) biomarkers. With BRAINGNNet at its core as the graph modeling framework, our proof-of-concepts and prototyping efforts integrate the brain topology with EEG data from public datasets, to differentiate between healthy and chronic pain subjects through Graph Neural Networks (GNNs).
Our initial findings suggest considerable potential for improving chronic pain monitoring and management in clinical applications.
Full Article
BRAINGNNet: Graph Models for Chronic Pain Detection through EEG Biomarkers
Motivation and Background
Chronic pain affects millions globally and is often accompanied by cognitive and emotional challenges, including mental disorders. The inability to objectively measure pain presents numerous issues in pain management, such as medication availability, clinical care, treatment monitoring, and clinical trials. Therefore, reliable, ethical, noninvasive detection and continuous measurement are crucial for effective treatment. Brain imaging, particularly EEG, shows promise due to its high temporal resolution and cost-effectiveness compared to fMRI and PET.
Recent EEG-based approaches to pain detection include univariate methods, which focus on individual metrics like location and magnitude of brain activation, and multivariate approaches leveraging machine learning (ML) to predict pain-related outcomes. These approaches, especially those using ML, require careful validation to ensure specificity and accuracy. Emerging efforts using deep learning (DL) frameworks on EEG data are expanding our understanding of pain as a dynamic brain network activity. However, no study has yet considered the topological relationships between EEG channels and band power activity in the context of pain detection using graph-based learners.
Biomarker Assessment Framework and Model Development
EEG Study Datasets
To evaluate the effectiveness of the proposed Heterogeneous Graph Neural Network (HGNN) learner, we developed a graph model aligning brain topology with the International 10-10 system for EEG electrode placement. We utilized datasets from two key studies: the cpCGX-BIDS (Chronic Pain Data EEG Dataset) and the MBB LEMON (Control EEG Dataset). We focused on the Eyes-Closed resting-state EEG reads to establish a baseline for chronic pain versus control subjects.
Preprocessing and EEG Feature Extraction
We employed DISCOVER-EEG, an open-source EEG pipeline, for preprocessing steps such as line noise removal, high-pass filtering, bad channel removal, and artifact detection. The feature extraction focused on Absolute Band Power (ABP) as a key EEG feature across frequency bands: Delta, Theta, Alpha, Beta, and Gamma.
Topology Mapping and Graph Model Development
We aligned the EEG channel locations with the International 10-20 system to create a Reference Brain Topology Map. This allowed us to compare activation paths and discover potential patterns. The graph database schema captures the brain topology and ABP marker nodes, representing EEG channel reads and clustering them around control versus chronic pain subjects.
GNN Architecture and Inference
The computational approach for the GNN involves using Heterogeneous GNN architectures, which are better suited for the proposed EEG biomarker approach. These architectures handle different types of nodes and edges and require specialized message-passing mechanisms. Candidate architectures include Multi-view Heterogeneous Graph Neural Networks and Heterogeneous Graph Attention Networks, which provide higher representational power and scalability.
Preliminary Analysis and Findings
Method
We performed a comparative analysis of EEG graphs for control and chronic pain subjects, focusing on variations in ABP across brain regions and hemispheres. We used Cypher to extract datasets from the Neo4J Graph DB to perform basic statistics and comparative analysis using Jupyter Notebooks.
Main Findings
We observed higher standard deviations in chronic pain subjects, indicating high-value variation and abnormal data distribution. Median values were more effective in distinguishing ABP intensities. There were consistent correlations between chronic pain and ABP in specific brain regions, notably the Alpha, Theta, and Gamma bands. Asymmetries in ABP favoring the right hemisphere emerged as potential differentiators for chronic pain.
Discussion: Alignment, Limitations, and Challenges
Our preliminary results align with other studies indicating increased Gamma, Theta, and Alpha powers in chronic pain patients. However, limitations with our analysis and findings include the chronic pain dataset’s heterogeneity, the lack of information covering participant characteristics, and the need for extended EEG datasets covering multiple sessions per participant to establish specificity and sensitivity.
Future Considerations
Our preliminary analysis supports the potential of BRAINGNNet in chronic pain detection using HGNNs. Future R&D streams include institutional collaborations for data acquisition, optimizing GNN models for low-resource environments, and developing real-time EEG processing pipelines with research-grade EEG headsets. These efforts aim to enhance chronic pain management and pave the way for advanced BCI consumer devices for clinical applications.
