Baseline deconvolution, phase correction, and signal quantification in fourier localized spectroscopic imaging

Abstract

Lineshapes of spectra obtained through chemical-shift imaging are often distorted due to the delay in sampling necessary for application of phase-encoding gradients. We have developed an automated fitting procedure which simultaneously performs signal quantification, phase correction, and baseline deconvolution of such spectra. The fit is based on the maximum likelihood method and can be implemented in either the time or the frequency domain.

High spatial resolution and speed in MRSI

“…Early experiments had phase encode delays of the order of 3-4 ms which produced significant baseline distortions. 33 Recent improvements in gradient performance have now reduced the delay to 0.5-1.0 ms. Figure 1 shows 31 P from a large region of brain tissue with simulated phase encode delays ranging from 0.5 to 3.5 ms. The appearance of the peaks are similar in each case, but the baseline of the spectrum depends upon the magnitude of the delay.…”

Pediatric Brain Tumors: Magnetic Resonance Spectroscopic Imaging

“…Typically, the data are transferred off-line to the remote Sun workstation, converted into a standard data format, fourier-transformed and phased using appropriate spectroscopic packages. To reliably and reproducibly quantify in vivo spectra, requires removal of baseline components, identification of peaks, and estimation of peak parameters, which can be accomplished using several different approaches (Barkhuijsen et al, 1985;Hore, 1985;Laue et al, 1985;Nelson and Brown, 1987;Spielman et al, 1988;Van der Veen et al, 1988;Derby et al, 1989;Nelson and Brown, 1989). Characteristics of the proton MR spectroscopy data that guide the choice of methodology are the larger number of spectra that need to be considered, and the need for whatever method is chosen to be robust to differences in signal to noise and peak configurations corresponding to different tissue types.…”

Automatic Estimation of the Time Origin and Overall Phase of an FID Signal

“…As for more complicated signals, one can easily verify that the absolute value of the sum of an arbitrary number [1] of undamped sinusoids with equal phases always reaches a maximum at t Å 0. In addition, the period of the constructive interference grows rapidly with the number of participating in which c k , a k , v k are the complex-valued amplitude, damping factor, and angular frequency of sinusoid k. Furthermore, t Å nDt / t 0 , where t 0 is the point of time of the first sample relative to the time origin, Dt is the sample interval, n is the sample index, n Å 0, .…”

“…Among frequency and the time domain by using iterative nonlinear other things, a delay t A is incurred. The phase of signal (b) least-squares (nlls) model-function fitting (1). An important at t Å t d / t A equals that of signal (a) at t Å t d .…”

“…We consider now the simple case of two sinusoids with parameters c 1 És(t)É Å 2ÉcÉexp(at)cos(Dvt/2).…”