Quantitative measurements of cerebral blood oxygen saturation using magnetic resonance imaging

Abstract

A quantitative estimate of cerebral blood oxygen saturation is of critical importance in the investigation of cerebrovascular disease because of the fact that it could potentially provide information on tissue viability in vivo. In the current study, a multi-echo gradient and spin echo magnetic resonance imaging sequence was used to acquire images from eight normal volunteer subjects. All images were acquired on a Siemens 1.5T Symphony whole-body scanner (Siemens, Erlangen, Germany). A theoretical signal model, which describes the signal dephasing phenomena in the presence of deoxyhemoglobin, was used for postprocessing of the acquired images and obtaining a quantitative measurement of cerebral blood oxygen saturation in vivo. With a region-of-interest analysis, a mean cerebral blood oxygen saturation of 58.4%+/-1.8% was obtained in the brain parenchyma from all volunteers. It is in excellent agreement with the known cerebral blood oxygen saturation under normal physiologic conditions in humans. Although further studies are needed to overcome some of the confounding factors affecting the estimates of cerebral blood oxygen saturation, these preliminary results are encouraging and should open a new avenue for the noninvasive investigation of cerebral oxygen metabolism under different pathophysiologic conditions using a magnetic resonance imaging approach.

FIG. 1

Sequence diagram of the multi-echo gradient and spin echo. G_SS, G_PE, and G_R represent the slice select, phase encoding, and frequency encoding, respectively. Analog-to-digital converter (ADC) indicates where images are acquired.

FIG. 2

Anatomic images from the first echo (A), spin echo (B) and the last echo (C) obtained from the multi-echo gradient and spin echo sequence are shown.

FIG. 3

Experimentally measured signal is plotted as a function of TE (A) and the measured magnetic resonance signal after removing the R2 effect (B). Unfilled squares indicate the echoes used for the R2 estimates. In contrast, the filled box indicates the echoes used for the R2′ estimates, whereas the unfilled box represents the echoes used to fit a second order polynomial function.

FIG. 4

R2 (A), R2* (B), R2′ (C), volume fraction (D), and color coded oxygen saturation Y (E) are shown. Color bar indicates the corresponding oxygen saturation from 0% to 100%.