Open AccessC ardiorespiratory noise correction improves the ASL signal
Author(s) -
Hassanpour Mahlega S.,
Luo Qingfei,
Simmons W. Kyle,
Feinstein Justin S.,
Paulus Martin P.,
Luh WenMing,
Bodurka Jerzy,
Khalsa Sahib S.
Publication year - 2018
Publication title -
human brain mapping
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.005
H-Index - 191
eISSN - 1097-0193
pISSN - 1065-9471
DOI - 10.1002/hbm.24000
Subject(s) - communication noise , resting state fmri , noise (video) , cerebral blood flow , cardiorespiratory fitness , signal to noise ratio (imaging) , signal (programming language) , neuroscience , computer science , mathematics , psychology , medicine , cardiology , artificial intelligence , statistics , philosophy , linguistics , image (mathematics) , programming language
Cardiorespiratory fluctuations such as changes in heart rate or respiration volume influence the temporal dynamics of cerebral blood flow (CBF) measurements during arterial spin labeling (ASL) fMRI. This “physiological noise” can confound estimates of resting state network activity, and it may lower the signal‐to‐noise ratio of ASL during task‐related experiments. In this study we examined several methods for minimizing the contributions of both synchronized and non‐synchronized physiological noise in ASL measures of CBF, by combining the RETROICOR approach with different linear deconvolution models. We evaluated the amount of variance in CBF that could be explained by each method during physiological rest, in both resting state and task performance conditions. To further demonstrate the feasibility of this approach, we induced low‐frequency cardiorespiratory deviations via peripheral adrenergic stimulation with isoproterenol, and determined how these fluctuations influenced CBF, before and after applying noise correction. By suppressing physiological noise, we observed substantial improvements in the signal‐to‐noise ratio at the individual and group activation levels. Our results suggest that variations in cardiac and respiratory parameters can account for a large proportion of the variance in resting and task‐based CBF, and indicate that regressing out these non‐neuronal signal variations improves the intrinsically low signal‐to‐noise ratio of ASL. This approach may help to better identify and control physiologically driven activations in ASL resting state and task‐based analyses.