We propose a method for correcting for bias
introduced by an iso-metabolic assumption in hypercapnia calibrated BOLD
studies. A graded hypercapnia design and an assumption of linear CMRO2
dependence on hypercapnia level are used to separate the calibration parameter M
from CMRO2 changes during hypercapnia. This method avoids intra-subject
and experimental variability introduced by making a prior assumption of iso-metabolism
or a CMRO2 decrease with hypercapnia based on literature values. We
implement this method using two distinct levels of hypercapnia, measuring lower
M values than when making the iso-metabolic assumption, with a significant
dose-wise reduction in CMRO2 with hypercapnia level.
