Cerebrovascular reactivity mapping using intermittent breath modulation

Cerebrovascular reactivity (CVR), an index of brain vessel's dilatory capacity, is typically measured using hypercapnic gas inhalation or breath-holding as a vasoactive challenge. However, these methods require considerable subject cooperation and could be challenging in clinical studies. More recently, there have been attempts to use resting-state BOLD data to map CVR by utilizing spontaneous changes in breathing pattern. However, in subjects who have small fluctuations in their spontaneous breathing pattern, the CVR results could be noisy and unreliable. In this study, we aim to develop a new method for CVR mapping that does not require gas-inhalation yet provides substantially higher sensitivity than resting-state CVR mapping. This new method is largely based on resting-state scan, but introduces intermittent modulation of breathing pattern in the subject to enhance fluctuations in their end-tidal CO2 (EtCO2) level. Here we examined the comfort level, sensitivity, and accuracy of this method in two studies. First, in 8 healthy young subjects, we developed the intermittent breath-modulation method using two different modulation frequencies, 6 seconds per breath and 12 seconds per breath, respectively, and compared the results to three existing CVR methods, specifically hypercapnic gas inhalation, breath-holding, and resting-state. Our results showed that the comfort level of the 6-second breath-modulation method was significantly higher than breath-holding (p=0.007) and CO2-inhalation (p=0.015) methods, while not different from the resting-state, i.e. free breathing method (p=0.52). When comparing the sensitivity of CVR methods, the breath-modulation methods revealed higher Z-statistics compared to the resting-state scan (p<0.008) and was comparable to breath-holding results. Next, we tested the feasibility of breath-modulation CVR mapping (6 seconds per breath) in 21 cognitively normal elderly participants and compared quantitative CVR values to that obtained with the CO2-inhalation method. Whole-brain CVR was found to be 0.150±0.055 and 0.154±0.032 %ΔBOLD/mmHg for the breath-modulation and CO2-inhalation method, respectively, with a significant correlation between them (y=0.97x, p=0.007). CVR mapping with intermittent breath modulation may be a useful method that combines the advantages of resting-state and CO2-inhalation based approaches.