Time-resolved MRI oximetry for quantifying CMRO(2) and vascular reactivity

Abstract

This brief review of magnetic resonance susceptometry summarizes the methods conceived in the authors' laboratory during the past several years. This article shows how venous oxygen saturation is quantified in large draining veins by field mapping and how this information, in concert with simultaneous measurement of cerebral blood flow, yields cerebral metabolic rate of oxygen, the brain's rate of oxygen consumption. The accuracy of this model-based approach in which the blood vessel is approximated as a long, straight cylinder, for which an analytical solution for the induced field exists, is discussed. It is shown that the approach is remarkably robust, allowing for time-resolved quantification of whole-brain metabolism at rest and in response to stimuli, thereby providing detailed information on cerebral physiology in health and disease not previously amenable by noninvasive methods.

Figure 1

Schematic showing relationship between applied field, B_o, vessel orientation and angles θ and φ in Eqs 2 and 3.

Figure 2

Measurement of venous oxygen saturation: a) magnitude gradient-echo image at the level of the common carotid arteries; phase-difference image showing increased phase for jugular veins (arrows) but not arteries. From the intra- to extravascular phase difference SvO₂ in the jugular vein was computed from Eq. 6 as 65%.

Figure 3

a) Axial gradient-echo image showing the superior sagittal sinus (SSS); (b) 3D rendition of SSS and orientation relative to the applied field; (c) analytically computed field map at location 2 (panel B); (d) model-based field map for the infinite cylinder approximation and the same tilt angle of 20.1°, yielding very similar field pattern, with an average intravascular field difference of 0.010 ppm, corresponding to 2.6% HbO₂ (from Li et al [26], with permission).

Figure 4

a) Correlation between AVO₂D and total cerebral blood flow in eight subjects (R²=0.74, P<0.006). Note that subjects with higher CBF values tend to have lower AVO₂D; b) Scatter plot of the cerebral metabolic rate of oxygen (CMRO₂) for each subject over three scanning sessions illustrating reproducibility. The vertical span of each diamond represents the 95% confidence interval for each group (from Jain et al [28], with permission).

Figure 5

Images from which venous oxygen saturation (SvO₂) and total cerebral blood (tCBF) flow during baseline, hypercapnia and recovery were derived: a) Axial magnitude image distal to the carotid bifurcation, used for quantification of tCBF; b) velocity images during baseline, hypercapnia and recovery showing increased velocity during the stimulus phase; c) axial magnitude image showing SSS; d) phase difference images during baseline, hypercapnia and recovery showing reduced intravascular phase during hypercapnia, commensurate with decreased oxygen extraction during hypercapnia (modified, from [39], with permission).

Figure 6

Time-course of tCBF and S_vO₂ during normocapnia (purple) and hypercapnia (pink) in a representative subject showing how the two parameters change in concert with each other. Note undershoot during recovery from hypercapnia (from [39], with permission).

Figure 7

a) Time-course plot of cohort-averaged S_aO₂, S_vO₂, and tCBF absolute parameter values; b) tCBF, AVO₂D, and CMRO₂ absolute parameter values. Error bars indicate ± 1 SD, shaded areas the apnea period. All values in time-course plots represent averages across the three repeated blocks of the paradigm. The bracketed sections ‘Base’ and ‘EA’ indicate the data used for computing average baseline values and end-apnea values (from Rodgers et al [43], with permission).