MR-visible water content in human brain: A proton MRS study

In vivo measurement of metabolite concentrations in the human brain by means of proton-MRS contributes significantly to the clinical evaluation of patients with diseases of the brain. The fully relaxed water signal has been proposed as an internal standard for calibration of the MRS measurements. The major drawbacks are the necesity to make the assumptions that the water concentration in the brain and that all tissue water is MR-visible. A number of in vivo measurements were carried out to estimate the concentration of MR-visible water in the brain of healthy volunteers divided into four age groups: newborn (0–23 days), adolescents (10–15 yr), adults (22–28 yr), and elderly people (60–74 yr). The examinations were carried out using a Siemens Helicon SP 63/84 MR-scanner operating at 1.5 T. Except for the newborn, four regions were studied in each subject using stimulated echo (STEAM) sequences without water suppression. In vitro measurements on a standard phantom were used for calibration. The calculated water concentrations ranged between 35.8 and 39.6 (mean 36.9) mol·kg wet weight]⁻¹ in the three groups, whereas it was 51.5 mol·kg wet weight]⁻¹ in the newborn, p<.01. The observed water concentration of neither the four regions nor of the three oldest age groups were significantly different. Comparisons between the water concentrations measured and those expected based on estimation of the content of grey and white matter in the region of interest from T₁-weighted images and biochemical data published, suggest that only a small fraction (<5%) of the tissue water may be MR-invisible. The study of healthy volunteers thus shows that errors introduced by using the unsaturated water signal for calibration are less than 10%, which is comparable to expected errors when other calibration procedures are used under similar measurement conditions.