Data Processing Methods in MRS - Advance Image Processing in Magnetic Resonance Imaging - page 405

800

700

NAA

Cr

600

500

Cho

400

a

NAA

300

200

100

0

4 3.5 3 2.5 2 1.5

p.p.m.

1 0.5 0 − 0.5 − 1

FIGURE 12.1 Quantification of N -acetyl-aspartate (NAA) concentration by manual inte-

gration of the absorption-mode H MRS spectral profile. The presence of baseline wan-

dering makes it difficult to define the area of integration. Also, coline (Cho) and creatine
(Cr) peaks are partially overlapped; in this case it is difficult to correctly define the
integration range.

1

area under the spectral peaks of interest is the oldest and simplest method used
to extract quantitative metabolite concentrations from the FID. In practice, this
approach is slightly inaccurate, and a bias is usually introduced by the operator
and baseline correction technique. As shown by Equation 12.3 and Equation 12.4,
the results rely on the correct phasing of the spectral components and depend on
the width of the integration area (ideally, from ; in practice, around the

spectral peak). Defining the integration width cannot be a trivial task, especially
when spectral peaks are partially superimposed or when acquisition artifacts or
broadband resonances (i.e., water signal contribution in

−∞ +∞

1

H-MRS) may distort the

spectrum baseline ( Figure 12.2a ). Acquisition noise and baseline waving may

also mask the contributions of low-amplitude metabolite components, whose

concentration is nonquantifiable ( Figure 12.3 ).

Therefore, in the last decade, a wide range of signal processing methods have
been proposed to obtain automatic, reliable quantification of metabolite concen-
trations from the recorded FID signals. Calculation of metabolite concentration
can be carried out in both time and frequency domain, requiring accurate esti-
mation of either the amplitude or the peak area.
This chapter will review the main signal processing methodologies used in
this field, and it will survey the successive processing steps that lead to the
calculation of molecular concentration of various metabolites contributing to the
recorded FID.