MELF: a multi-view ensemble learning framework for normative resting state EEG signal quality assessment

Clinical Snapshot

Key Findings

• P Value: Not reported

• Effect Size: MELF achieved 97.69% accuracy (AUC = 0.9957) on internal validation across four open datasets, and 92% accuracy (AUC = 0.88) on an independent expert-annotated clinical test set; up to 10% accuracy improvement over single-view methods

• Primary Outcome: Classification accuracy for normative rsEEG signal quality (acceptable vs. unacceptable)

• Nnt Or Sensitivity: Sensitivity and specificity not reported; AUC = 0.9957 (internal), AUC = 0.88 (independent test set). Note: the drop from AUC 0.9957 to 0.88 on independent testing warrants careful interpretation regarding overfitting

• Confidence Interval: Not reported

Clinical Application

The framework uses open-source feature extraction methods (ICLabel, phase lag index, spectral analysis) and a random forest classifier, suggesting feasibility of implementation in existing EEG processing pipelines. Computational efficiency advantage over deep learning is noted. However, practical deployment requires software packaging, validation on local EEG systems, and integration with existing preprocessing workflows. In Australia, rsEEG is increasingly used in research contexts for psychiatric disorder biomarker development (e.g., depression, schizophrenia) and in academic neuroscience. The RACGP does not currently provide specific guidelines for rsEEG quality control in primary care. The TGA does not regulate EEG signal processing software as a therapeutic good unless it constitutes a medical device with diagnostic claims. Australian neuroscience research groups using large normative EEG datasets (e.g., through the Australian Imaging Biomarkers and Lifestyle study infrastructure) could benefit from automated quality control tools. PBS listing is not applicable. Adoption would require validation against Australian normative EEG datasets and local expert annotation standards. Researchers and clinical neurophysiologists performing normative resting-state EEG acquisition in healthy populations for neuroscience research or psychiatric biomarker development; potential extension to clinical EEG quality control pipelines

Abstract

Normative resting-state electroencephalography (rsEEG), defined as standardized recordings from healthy populations, provides essential baselines for studying brain mechanisms and psychiatric disorders. However, its reliability critically depends on rigorous quality assessment, as artifacts or over-processing can distort subsequent analyses. Current approaches often rely on single-feature perspectives, which fail to capture the multidimensional nature of signal quality. We propose MELF, a quality assessment framework for normative rsEEG that systematically integrates three complementary feature domains: time domain (mobility, complexity), frequency domain (-band relative power, parallel log spectra index (PaLOSi)), and spatial domain (-band directed phase lag index, bad channel ratio, ICLabel-classified brain component proportion). A random forest-based ensemble classifier is constructed using dynamic weight allocation based on cross-validation accuracy (Acc). Experiments on four open rsEEG datasets (DSTR, TRCS, LEMON, and HBN) and expert-annotated clinical data demonstrate that: (1) MELF achieves up to 10% Acc improvement over single-view methods; (2) MELF outperforms traditional machine learning and deep learning baselines while reducing computational time, reaching 97.69% Acc (AUC = 0.9957), and maintains strong performance on an independent test set (92% Acc, AUC = 0.88); (3) the spectral view contributes the most, confirming the evaluative capability of the PaLOSi. MELF provides an interpretable, high-Acc, and robust automated framework for normative rsEEG quality assessment, with broad applications in neuroscience research and clinical quality control.

References

1. 1.Lv, W., Jia, D., Zha, Z., Zhou, B., Lv, Z., Valdes-Sosa, P. A., & Hu, S. (2026). MELF: a multi-view ensemble learning framework for normative resting state EEG signal quality assessment. Biomedical Physics & Engineering Express. https://doi.org/10.1088/2057-1976/ae84d2