Measurement ofR′2in the Presence of Multiple Spectral Components Using Reference Spectrum Deconvolution
✍ Scribed by F.W. Wehrli; J. Ma; J.A. Hopkins; H.K. Song
- Publisher
- Elsevier Science
- Year
- 1998
- Tongue
- English
- Weight
- 464 KB
- Volume
- 131
- Category
- Article
- ISSN
- 1090-7807
No coin nor oath required. For personal study only.
✦ Synopsis
A method is described for measuring R 2 , the RF reversible contrithe different R 2 (Å1/T 2 ) relaxation rates of the various probution to the effective transverse relaxation rate in yellow trabecular tons in FATG, which can cause errors in R 2 obtained by marrow, as a means to evaluate trabecular bone structure and denmeans of the GESFIDE technique (7). For example, the sity. The method exploits the similarity in spectral composition of methylene protons a to the carboxyl group experience less the marrow and fat in subcutaneous tissue. Under these conditions segmental motion and therefore relax faster, thus causing the gradient echo envelope of the marrow signal can be regarded multiexponential R 2 decay. Further, some of the modulations as a convolution of a function describing the bone marrow intravoxel have long periods (on the order of a typical sampling period line broadening (R 2 ) with a function expressing chemical shift modof the interferogram); therefore the apparent R * 2 and R 2 ulation, which is obtained from the echo envelope of the subcutanevalues will depend on the duration of the echo train. ous fat signal in a reference region. Simple division of each of a In this paper a simple method for extracting R 2 is proposed series of echoes by the reference signal is shown to afford a smooth decay which can be fitted to a model to extract R 2 . The method that effectively removes the signal modulation by deconvoluhas been evaluated in the upper femur of test subjects and a strong tion with a reference signal collected in the same scan from correlation of the thus derived R 2 values with those obtained by adipose tissue in which line broadening due to local susceptithe GESFIDE technique is demonstrated. The close correspondence bility variations is negligible. Unlike GESFIDE (gradientin spectral composition of proximal femur marrow and subcutaneecho sampling of FID and echo), this method thus circumous fat is further illustrated by means of localized spectroscopy. vents a measurement of R 2 , and its only requirement is that The major potential error source is global inhomogeneity in the the chemical composition of the tissue used for the reference reference region which can lead to an underestimation of the designal equals that in bone marrow and that global inhomogemodulation-derived R 2 .