Riemannian Geometry Applied to Detection of Respiratory States From EEG Signals: The Basis for a Brain–Ventilator Interface

• Published in: IEEE transactions on biomedical engineering Vol. 64; no. 5; pp. 1138 - 1148
• Main Authors: Navarro-Sune, X., Hudson, A. L., Fallani, F. De Vico, Martinerie, J., Witon, A., Pouget, P., Raux, M., Similowski, T., Chavez, M.
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.05.2017; The Institute of Electrical and Electronics Engineers, Inc. (IEEE); Institute of Electrical and Electronics Engineers
• Subjects: Adult; Air flow; biomedical monitoring; Biomedical signal processing; Brain; Brain - physiology; Brain-Computer Interfaces; brain–computer interfaces (BCI); Breathing; Classification; Computer applications; Computer Science; Cortex; Covariance matrices; Covariance matrix; Diagnosis, Computer-Assisted - methods; EEG; Electrodes; Electroencephalography; electroencephalography (EEG); Electroencephalography - methods; Feature extraction; Female; Geometry; Human-Computer Interaction; Human-computer interface; Humans; Implants; Iterative methods; Machine Learning; Male; Mechanical ventilation; medical signal detection; Pattern Recognition, Automated - methods; Prototypes; Respiration; Respiration, Artificial - methods; Respiratory Mechanics - physiology; Synchronization; Ventilation; Ventilators; Verbal communication
• ISSN: 0018-9294; 1558-2531; 1558-2531
• DOI: 10.1109/TBME.2016.2592820

Goal: During mechanical ventilation, patient-ventilator disharmony is frequently observed and may result in increased breathing effort, compromising the patient's comfort and recovery. This circumstance requires clinical intervention and becomes challenging when verbal communication is difficult. In this study, we propose a brain-computer interface (BCI) to automatically and noninvasively detect patient-ventilator disharmony from electroencephalographic (EEG) signals: a brain-ventilator interface (BVI). Methods: Our framework exploits the cortical activation provoked by the inspiratory compensation when the subject and the ventilator are desynchronized. Use of a one-class approach and Riemannian geometry of EEG covariance matrices allows effective classification of respiratory states. The BVI is validated on nine healthy subjects that performed different respiratory tasks that mimic a patient-ventilator disharmony. Results: Classification performances, in terms of areas under receiver operating characteristic curves, are significantly improved using EEG signals compared to detection based on air flow. Reduction in the number of electrodes that can achieve discrimination can be often desirable (e.g., for portable BCI systems). By using an iterative channel selection technique, the common highest order ranking, we find that a reduced set of electrodes (n = 6) can slightly improve for an intrasubject configuration, and it still provides fairly good performances for a general intersubject setting. Conclusion: Results support the discriminant capacity of our approach to identify anomalous respiratory states, by learning from a training set containing only normal respiratory epochs. Significance: The proposed framework opens the door to BVIs for monitoring patient's breathing comfort and adapting ventilator parameters to patient respiratory needs.