A linear model for event-related respiration responses

- We develop a novel method for analysing event-related respiratory responses.
- This method is based on a Psychophysiological Model (PsPM) of interpolated time series.
- We analyse respiration period (RP), amplitude (RA) and flow rate (RFR).
- RA and RFR estimates distinguish different event types, and all three measures distinguish events from non-events.
- The new method could be useful for fMRI experiments using respiration belts.

BackgroundCognitive processes influence respiratory physiology. This may allow inferring cognitive states from measured respiration. Here, we take a first step towards this goal and investigate whether event-related respiratory responses can be identified, and whether they are accessible to a model-based approach.New methodWe regard respiratory responses as the output of a linear time invariant system that receives brief inputs after psychological events. We derive average responses to visual targets, aversive stimulation, and viewing of arousing pictures, in interpolated respiration period (RP), respiration amplitude (RA), and respiratory flow rate (RFR). We then base a Psychophysiological Model (PsPM) on these averaged event-related responses. The PsPM is inverted to yield estimates of cognitive input into the respiratory system. This method is validated in an independent data set.ResultsAll three measures show event-related responses, which are captured as non-zero response amplitudes in the PsPM. Amplitude estimates for RA and RFR distinguish between picture viewing and the other tasks. This pattern is replicated in the validation experiment.Comparison with existing methodsExisting respiratory measures are based on relatively short time-intervals after an event while the new method is based on the entire duration of respiratory responses.ConclusionOur findings suggest that interpolated respiratory measures show replicable event-related response patterns. PsPM inversion is a suitable approach to analysing these patterns, with a potential to infer cognitive processes from respiration.