Kevin Murphy1, Rami K. Niazy1,2, C. John Evans1, Richard G. Wise1
1Cardiff University Brain Research Imaging Centre, School of Psychology, Cardiff University, Cardiff, UK; 2Biomedical Physics Department, King Faisal Specialist Hospital & Research Centre, Riyadh, Saudi Arabia
Functional connectivity analysis exploits coherent low-frequency oscillations in the BOLD signal. Fluctuations in arterial CO2 are low-frequency and can confound interpretation of neuronal activity. Two correction methods that attempt to remove this noise source use end-tidal CO2 and RVT measures to regress breathing-related changes from the BOLD time series. We show that a lack of similarity between the CO2 and RVT regressors and the variance explained by them implies that each is measuring a different portion of the noise related to fluctuations in breathing and arterial CO2. This suggests that both measurements are required for accurate removal of such noise.