TARQUIN analysis

MRS Software Tarquin
1

Dear All,

Hi

What is the metabolite TARQUIN is considering for SNR and FWHM estimation?
I mean which of the main peaks?

Thank you

2

I’m not sure how recent the manual is but there are some definitions here: TARQUIN User Guide Version 4.3.5

3

Hi,

Thank you very much.But it’s not specified which metabolites are the criteria for these parameters…

4

Hi NNeda,

as far as the Guide says, the SNR is calculated from the maximum of the spectrum - the baseline in the 0.5 to 4 ppm region which usually is the tNAA peak. The SNR max metric does the same across the whole spectral range, therefore, including the water peak. And SNR metab is the ratio is the maximum of the spectrum - baseline divided by 2xRMS of the residual (between 1.9 to 4 ppm) which is a signal to residual ratio similar to the LCModel SNR output.

For the FWHM output Tarquin reports the FWHM value that is derived from the cross-correlation analysis for the frequency referencing including the major singlets tNAA, tCt, and tCho (see https://groups.google.com/g/tarquin_users_group/c/glFhrvNer68/m/ORnPxfYpCgAJ).

Best,
Helge

5

Dear Helge,
Hi

Thank you very much for your kind attention :rose: :pray:

6

Dear All,

Why is Cr considered as a standard for metabolite FWHM?

Thank you

7

which one is considered mainly as noise of spectrum? spectral noise or time domain noise?

8

Cr is a strong singlet that is present in most scenarios - one could use NAA, too, but it can be diminished substantially in tumors, for example.

9

I’m not sure I understand the question - considered by who? The MRS signal is, per se, acquired as a time-domain signal, and noise represents random fluctuations due to thermal variations in the sample and the receiver hardware. However, many software solutions (but probably not all?!) calculate the noise from a de-trended signal-free area of the frequency-domain spectrum, if that helps.

See the MRS terminology paper for some introductory language.

10

In comparison between two spectra I mean…

11

I mean, you can compare the noise of two spectra from their time-domain noise or their frequency-domain noise. It doesn’t matter. FFT is a linear operation.

12

Thank you so much. :rose:

13

How do these ratios: tCho/tNAA, tCho/NAA and Cho/NAA differ from each other in tumor assessment?