| Popis: |
Quantitative information on the fat and water contents of tissue is important for biomedical investigations. For example, it allows more accurate diagnosis and evaluation of treatment in a variety of liver and bone marrow diseases, such as fatty liver degeneration, cirrhosis, hepatitis, malignant liver infiltration or disorders of the hematopoiesis, malignant and inflammatory bone marrow infiltration, and aseptic bone necrosis. Proton chemical-shift imaging is the preferred method to measure the distribution of fat and water in tissue. So far, however, the method has been used mainly to qualitatively enhance tissue contrast, rather than to obtain quantitative data. Recently, Buxton et al. (I) demonstrated with phantom investigations, that, employing the Dixon measuring sequence (2), one is well able to measure fat and water contents with an accuracy of better than 5%. In essence, their data evaluation computes the difference of the magnitude images of the in-phase echo (simultaneous gradient echo and Hahn spin echo (3)) and the out-of-phase echo (Hahn echo delayed with respect to the gradient echo to dephase the fat and water magnetizations by 180”) to deduce the tissue composition. The advantage with this data processing is the accurate information on the absolute value of the difference between fat and water contents. On the other hand, the method is not able to determine which one of the compounds is more abundant. In this note, we propose a method, which, from a conventional Dixon data set and without operator interaction, shows which one of the two compounds fat and water is more abundant, and therefore, in combination with the Buxton data evaluation, allows an unambiguous quantitation of tissue composition. The chemical-shift image data are acquired with a slightly modified Dixon sequence (2) as shown in Fig. 1. The slice-selection refocusing gradient is applied later in the sequence to act at the same time as a spoiler gradient for residual transverse magnetization generated by an improperly set composite refocusing RF pulse. In experiment (a), Hahn echo and gradient echo occur at the same time, whereas in experiment (b), the refocusing RF pulse is delayed by a time interval d such that at the time of the gradient echo, fat and water magnetizations are mutually 180” out of phase. Prior to the data acquisition, the B. homogeneity is optimized in the cross section of interest |
