Joint Seminar

• Centre for Neurotechnology
• Biomathematics Group
• EPSRC Centre for Mathematics of Precision Healthcare

Speaker: Dr Mario Chavez (The French National Centre for Scientific Research)

Title: Differential Geometry Applied to Monitoring of Brain States from EEG Signals

Abstract: Current neuroscience research attempts to understand how brain functions result from dynamic interactions in large-scale cortical networks, and to further identify how cognitive tasks or brain diseases contribute to reshape this organization. In this talk I’ll present method based on differential (Riemannian) geometry to identify a cortical signature of breathing discomfort from EEG recordings of patients. I’ll show how the characterization of spatio-temporal patterns on differential manifolds may provide much better performances than alternate methods currently used in brain computer interfaces. Further, I’ll show the effective translation of our algorithm to an embedded device (portable, noninvasive, with few electrodes, and fast computation) that can be highly operable in clinical environments as well as in custom-designed systems.

Refs:

X. Navarro-Sine et al. “Riemannian Geometry Applied to Detection of Respiratory States From EEG Signals: The Basis for a Brain–Ventilator Interface. IEEE Trans. Biomed. Eng., vol. 64, no. 5, pp. 920–928, 2012.

A. L. Hudson et al., “Electroencephalographic detection of respiratory-related cortical activity in humans: From event-related approaches to continuous connectivity evaluation” J. Neurophysiol., vol. 115, pp. 2214– 2223, 2016.

T. Similowski et al. “Method for characterising the physiological state of a patient from the analysis of the cerebral electrical activity of said patient, and monitoring device applying said method”. Patent WO2013164462A1