Measurement of tissue oxygen extraction ratios from venous blood T₂: Increased precision and validation of principle

Abstract

(1) have shown that this oxygenation-dependent contrast affects both T 2 and T * 2 in blood, while extravascular effects are predominantly related to T * 2 . In addition, it was recently demonstrated that macrovascular signal changes dominate in both gradient-echo (GRE) and spin-echo (SE) BOLD experiments at low (4 -6) and intermediate (7,8) field strengths. It has been suggested that this situation can be exploited as a means to directly determine venous blood oxygenation by measuring intravascular T 2 (9,10), T * 2 (4,11,12), or signal phase (13-15). These relaxation approaches find their origin in the literature data for pure blood, showing a direct relationship between oxygenation and transverse relaxation times (16 -21). We recently extended the well-known Luz-Meiboom equation (17,22,23), often used to describe blood relaxation, by expressing the erythrocyte relaxation rate (R 2,ery ) and the plasma-erythrocyte susceptibility shift difference in terms of the fractional concentration of deoxyhemoglobin, x deoxy (24,25). For venous blood, this deoxygenation fraction is determined by the oxygen saturation fraction (Y v ), which is a function of the arterial oxygen saturation fraction (Y a ) and the oxygen extraction ratio (OER) of the tissue from which the venous blood is draining (24,25):This OER, also called the oxygen extraction fraction (OEF), describes the balance between oxygen delivery and consumption in a tissue:in which [Hb tot ] is the total hemoglobin concentration in millimolar, CMR O2 the cerebral metabolic rate of oxygen in mol/g/min, and CBF the cerebral blood flow in ml/g/min. Thus, at constant hematocrit and arterial oxygenation, OER depends on the ratio of CMR O2 and CBF. Under these conditions, the intravascular venous BOLD effect on any physiological alteration (i.e., brain activation, hypoxia, hypercarbia) directly reflects the mismatch of changes in oxygen metabolism and blood flow (25). Using the theory outlined in the Materials and Methods section, we recently quantified absolute OER values in the human visual cortex during baseline activity and visual stimulation using T 2 measurements inside veins draining from the parenchyma (25). Quantification of absolute T 2 s was achieved from the echo time dependence of the venous signal intensity obtained using single spin-echo (SE) experiments. However, these so-called Hahn spin-echo experiments are time-consuming and have several disadvantages. First of all, the theory predicts that the apparent T 2 values measured with this kind of sequence are TE-dependent, and quantification of OER in the previous article could only be achieved by minimizing the difference between exponential signal-intensity fits through experimen-

Relationship of apparent myocardial T2 and oxygenation: Towards quantification of myocardial oxygen extraction fraction

“…As an example, at ϭ 6 msec, Hct ϭ 0.4, and OEF ϭ 0.39, the calculated blood T 2 by Eq. [2] is 165 msec, which agrees favorably with other reported values of 172 msec (16) and 159 msec (17), measured at the same conditions. A similar finding can be also obtained at ϭ 12 msec.…”

“…The constant values (susceptibility shifts, exchange lifetime, etc.) for deriving these parameters (13) were obtained through experiments on bovine blood at 1.5 T and 37°C (17). There was no information available on the errors associated with these constant values and thereby we could not determine the errors of blood T 2 by Eq.…”

“…In general, T 2b in the blood is a parabolic function of oxygen saturation fraction Y b or OEF as follows (16,17):…”

“…[7] were taken from those obtained in our experiments in dogs at 1.5 T, whereas parameters in Eq. [4] were calculated based on data in the literature (17). The R 20t was 22.8 Ϯ 1.4 seconds -1 and R 21t was 10.4 Ϯ 7.2 (ϫ 10 6 ) seconds -3 .…”

Inverse T2 contrast at 1.5 Tesla between gray matter and white matter in the occipital lobe of normal adult human brain

“…As shown in Table 2, such an increased OER will significantly shorten T 2 of capillary and venular blood. For instance, it can be calculated that parenchymal T 2 decreases by 0.6 msec when OER of occipital GM increases from 0.4 to 0.5 (19), thus further reversing the gray/white T 2 contrast pattern in the occipital lobe. However, this effect is very small and cannot be the main cause of the inverse contrast.…”