Digital Signal Processing Reference
In-Depth Information
reconstructed.
puted metabolite concentrations of the four mobile lipid components are also
presented therein. It can be clearly seen on the bottom left panel (iii) in
Fig. 8.2 that there are four components, whereas only two can be appreciated
from the total shape spectrum. This unambiguous quantitative information
extracted by the FPT via parametric analysis could be used to develop norma
tive data bases for mobile lipid concentrations in the normal brain versus the
corresponding findings seen in brain tumors, in order to provide needed stan
dards to aid in neurodiagnostics, identifying gliomas of various grades versus
nonmalignant pathology with specific patterns of departures from normal
concentrations of these lipid components.
8.3.4 Pade reconstruction of the components of total choline
at 3.2 ppm to 3.3 ppm on the MR brain spectrum
As seen in Fig. 8.2 in the spectral region between 3.2 ppm and 3.3 ppm
there are two closely lying peaks of choline and of phosphocholine, at 3.239
ppm and 3.301 ppm. On the absorption total shape spectra on the bottom
right panel in Fig. 8.2, these appear as a single peak. In contrast, these
two resonances are clearly distinguished on the absorption component spectra
(bottom left panel in Fig. 8.2). The importance for oncologic diagnostics
of identifying and exactly quantifying the components of choline has been
repeatedly emphasized, not only within the realm of neurooncology [268,
301, 302], but also for diagnostics within breast, prostate and other cancers
[32, 33], [307]-[310].
8.3.5 Pade reconstruction in the region between 3.6 ppm and
4.0 ppm on the MR brain spectrum
A similar approach can be also helpful for other regions of the spectrum
in which there have been major, clinicallyimportant dilemmas in neuro
diagnostics. These regions include 3.6 ppm to 4.0 ppm, where, as discussed,
there has been uncertainty as to the number of components. Here, the FPT
was shown to precisely identify the three resonances, assigned as glutamine
3.803 ppm, creatine 3.944 ppm and phosphocreatine (PCr) 3.965 ppm. This
was achieved even though PCr was completely underlying creatine, as seen on
the absorption component shape spectrum on the bottom left panel in Fig.
8.2. Once again, from the total shape spectrum these two resonances cannot
be distinguished (bottom right panel of Fig. 8.2).
Overall, it is remarkable that thus far MRS and MRSI have made tremen
dous strides in neurodiagnostics by relying merely upon a handful of metabo
lites or their ratios. We can anticipate that Padeoptimized MRS could indeed
be of vital important to further strides.
Search WWH ::
Custom Search