Motor Intention Quantization for Patients With Disorders of Consciousness by Multimodal BCI Combining
Nan Wang1,2,3, Xiaoke Chai2,3, Jiuxiang Song4
1Department of Neurosurgery, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China.
CNS Neuroscience & Therapeutics
|December 6, 2025
Summary
Multimodal electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) fusion improves the diagnosis of disorders of consciousness (DoC). This combined approach offers superior accuracy compared to single-modality techniques for assessing patient consciousness levels.
Area of Science:
- Neuroscience
- Medical Technology
- Clinical Diagnostics
Background:
- Single-modality electroencephalography (EEG) or functional near-infrared spectroscopy (fNIRS) show limitations in accurately assessing consciousness in patients with disorders of consciousness (DoC).
- Suboptimal diagnostic accuracy of unimodal approaches necessitates the development of more effective diagnostic tools for DoC.
- Clinical Trial Registry: ChiCTR2400085830.
Purpose of the Study:
- To investigate the efficacy of multimodal fusion technology combining EEG and fNIRS for the clinical diagnosis of DoC patients.
- To enhance the diagnostic accuracy for disorders of consciousness by integrating neurophysiological and hemodynamic data.
Main Methods:
- Eleven patients with DoC (six minimally conscious state [MCS], five vegetative state [VS]) were enrolled.
- Simultaneous EEG and fNIRS recordings were acquired using a motor intention-based brain-computer interface (MI-BCI) paradigm.
- Analysis included time-frequency analysis, event-related desynchronization (ERD), and hemodynamic changes (HbO, HbR, HbT), followed by multimodal classification.
Main Results:
- The EEG-fNIRS multimodal approach significantly outperformed unimodal EEG and standalone fNIRS in classifying healthy controls (HC), MCS, and VS patients (AUC 0.69 ± 0.10).
- Unimodal EEG achieved an AUC of 0.43 ± 0.19 (p < 0.01), and standalone fNIRS achieved 0.63 ± 0.10 (p < 0.05).
- The EEG_ERD index differentiated MCS from VS, with fNIRS_ACC, EEG_ACC, and fNIRS_slope being key indicators for classification.
Conclusions:
- Integrating multimodal MI-BCI paradigms shows significant clinical potential for evaluating consciousness levels in DoC patients.
- The synergistic combination of neurophysiological (EEG) and hemodynamic (fNIRS) biomarkers offers a robust framework for improving diagnostic precision.
- This multimodal strategy enhances the accuracy of bedside diagnostic protocols for disorders of consciousness.
