Biomedical subjects

T Raidy

Publications and source records attributed to T Raidy.

4 recordsLinked to original sources

PubMed1994

Automated single-voxel proton MRS: technical development and multisite verification.

To improve clinical utility, an integrated method has been developed to automatically acquire and process single-voxel in vivo proton spectra on a 1.5 T clinical scanner. This method includes automated adjustment of linear shims using a very rapid modified simplex method, automated water suppression, and applies a water referencing scheme to correct for phase and residual eddy current effects. No operator intervention is required for the acquisition and processing of these pure-absorption spectra. This method was tested in a preliminary multisite trial to determine intersite and intrasite variability of metabolite ratio measurements. In a sample of over 100 examinations, the standard deviation of the ratios NAA:Cr, Cho:Cr, and ml:Cr were found to be under 15% when using this method, a substantially narrower range than has been found in studies relying on manual adjustment of the instrument and/or manual processing. This result indicates that automated setting of acquisition and processing parameters is of critical importance in the clinical application of in vivo spectroscopy.

Algorithms↗

PubMed1994

Evaluation of the clinical performance of automated proton magnetic resonance spectroscopy in children.

RATIONALE AND OBJECTIVES: Because expeditious neuroimaging is imperative in pediatric patients, we evaluated automated procedures for proton magnetic resonance spectroscopy (1H MRS) of the brain of children. METHODS: 1H MRS was performed on a 1.5-T GE Signa. The protocol included stimulated echo-acquisition mode and spin-echo point resolved spectroscopy. The automated routine included adjustment of first-order gradient shims (x, y, z1) to optimize magnetic field homogeneity, transmit power, center frequency, receiver gain, and water suppression. All spectra were processed with the use of spectroscopy analysis software from General Electric on a Sun workstation. RESULTS: The use of the automated procedures reduced the length of our 1H MRS protocol by 50%. Magnetic field homogeneity was within our accepted standards (7 +/- 2 Hz). Water suppression was within range of our accepted factors (1000-10,000). However, on certain occasions, baseline distortions affected resonances in the 3.22-4.04 ppm range. CONCLUSIONS: Shortening of the time required for clinical 1H MRS will increase its application in evaluating children.

Brain↗

PubMed1993

Correction of phase effects produced by eddy currents in solvent suppressed 1H-CSI.

Accurate phasing of MRS spectra is often difficult unless time varying phase effects produced by gradient-induced eddy currents that persist during data acquisition are eliminated. This effect is particularly problematic in 1H-CSI spectra where frequency shifts produced by static field inhomogeneity and phase shifts produced by eddy currents combine. In this paper we present a method that corrects both shifts and eliminates manual phasing of individual CSI spectra typically required to recover a pure absorption line shape. The method uses a time domain phase correction derived from the ambient water signal acquired under identical conditions (i.e., acquisition parameters, gradient sequence) as the solvent-suppressed CSI data. Results from CSI experiments on phantoms and in vivo solvent suppressed 1H-CSI spectra from normal human brain are presented demonstrating the capabilities of the technique.

Algorithms↗

PubMed1993

Integrated 3D display of brain surface anatomy and MR spectral data.

We describe a method of displaying the relationship between MR spectral data and an MRI-derived three-dimensional (3D) model of the same subject's brain surface. In this way, biochemical abnormalities in the MR spectra can be localized with respect to specific gyral convolutions (e.g., those associated with movement and sensation), which are best identified on 3D brain models. This was accomplished by retrospectively registering spectral data and MR images, acquired with different head coils. The highly resolved MR images were used to identify the brain surface in the poorly resolved spectral data and to produce a 3D rendition of brain surface metabolite distributions. This was then integrated with an MRI-derived 3D model of brain gyral anatomy. The method was tested on P31 spectral data from a phantom and from a human volunteer.

Adult↗