EEG time-frequency maps during respiratory-related cortical activation in humans

Introduction Event-locked averaging of EEG recordings has demonstrated that preinspiratory potentials precede inspiration during respiratory-related cortical activation. However, this procedure relies on assumptions of on-going brain activity and is highly susceptible to low frequency artifacts. Frequency analysis is an alternative method of assessing EEG activity, but it is not known if this method can identify respiratory-related cortical activation. The aim of this study was to compare time-frequency maps during different respiratory 'tasks'. Methods Healthy subjects (n=6) performed 3 conditions: quiet breathing (QB), self-paced voluntary sniffs and ventilation with an inspiratory threshold load (ITL; ∼23 cmH20, range 18-25 cmH20). EEG recordings were made from 32 channels. Time-frequency maps were computed for each subject and condition for recordings from Fz, FCz, Cz, C3 and C4. Contrasts between conditions were performed by T score calculation with statistical analysis by clustering permutation. Results There were differences in cortical activation (alpha band, 8-13Hz) prior to the onset of inspiratory flow on Fz, C3 and C4 between QB and sniffs (p< 0.1) and on C4 between QB and ITL (p< 0.1). Conclusions Time-frequency maps varied between different respiratory conditions which suggests they can discriminate differences in cortical activation related to the control of respiration. Furthermore, we can identify the specific frequency components, in addition to time-locked changes in EEG. Analysis of EEG activity above 5Hz may lessen the effect of artifacts on physiological interpretation and increase the feasibility of EEG recordings to assess cortical activation in the clinical setting.