Friday 21 May 2010

Bruker Smiles

The above title is not aimed to mislead and it does not refer to Bruker’s state of mind: Quite simply, the purpose of this post is merely technical and the relation between Bruker and its smiles will be apparent in a moment… Just keep reading…

Because it wouldn't make sense otherwise, NMR instruments use receiver systems equipped with digital filters since a relatively long time ago. The advantages of such digital filters (generally designed as low-pass filters and applied together with oversampling and decimation methods) are many fold, ranging from higher quality spectral baselines to SNR and effective dynamic range improvements, enhanced reduction of potential sources of folded signals, etc

It´s not all about advantages though … I’m sure most of you are already very well aware of the pesky problem that is infamously known as group-delay artifact in Bruker (and Jeol) data which has plagued the NMR community since these companies switched to digital receivers. In short, the FID resulting from the digital filter does not start at time = 0 but only after a long and slowly rising oscillation of length G (G = Group Delay).

Some empirical procedures to correct it were presented on the internet but they are palliative and do not resolve the problem completely, particularly when apodization is applied.
Typically and depending on how the FID is processed, the spectrum might exhibit smiles (baseline artifacts pointing up) or frowns (baseline artifacts pointing down) at the outer regions of the spectrum as depicted below:

The ultimate solution

These small artifacts are in general not a big problem as one could use a spectral width large enough so that the peaks of interest in the spectrum will not be affected by these artifacts (although some processing algorithms such as backward Linear Prediction could be somewhat problematic with the Group Delay). In any case, we did not feel very comfortable with present solutions to this problem. A few months ago, I went for dinner with Stan and right after it, the power of the red wine and above all, the Galician octopus inspired Stan in such a way that he managed to understand the engineering drawback and proposed a new correction algorithm which we implemented together in Mnova just a few minutes later (whilst still under the influence of the wine :-) ).
Basically we have now a new pre-processing algorithm that corrects in a totally automatic way any Bruker FID corrupted by the group-delay artifact, producing a normal and physically correct FID so that the smiles will not be seen in the f-domain spectrum. The performance of the new algorithm is illustrated in the figure below:

This enhanced correction is available in Mnova since version 6.1.1 onwards, although it is not the default processing method for the moment. In order to activate it, it is necessary to select it via Processing/Group Delay menu command.

I guess the take home from the story is never underestimate the power of red wine and Galician octopus :-)


Anonymous said...

Varian spectrum does not have the baseline problem. It seems the spectrometer does this correction before tranfering data to host computer.

I'm interested the method used by MNova. Can you describe more? Thanks.

stan said...

Yes, the power of the Galician octopus is great! I hope the beast is not close to extinction, despite my voracity.

More seriously, and partially as an answer to "Anonymous":

Ever since having found out exactly what did Bruker and Jeol engineers wrong, I am planning to write an online article about the dangers of trying to slavishly emulate in digital way what used to be done before employing analog circuitry. By the way, this is an excellent electronics-textbooks example, since both groups made independently the same error (just a parameter differs). And yes, Varian neatly avoided the trap (don't ask me whether by genius or by sheer luck).

While the article will appear on my website (in 'Stan's Library'), me and Carlos plan to make available a freeware utility correcting the Bruker and Jeol FID's. A very simple one - just bad FID goes in, good FID goes out.

The planned timing for this little project (article plus utility) is second half of September, synchronized with the SMASH meeting in Portland. Needless to say, Mnova Users will not need the utility since Mnova already takes care of the problem.

Should Bruker or Jeol top brass want to know the trick in advance and/or do something about it, please contact Mestrelab (Carlos or Santi). They detain the rights to it and represent my interests, too.

Anonymous said...

Thanks a lot Carlos, looking forward at seeing the detailed solution.

Axel Meissner said...

Dear Carlos, the group delay discussion has been around for quite a while, probably because the digital filter methods were not that well documented in the beginning but since this information is available and the group delay value is actually included in the status parameters - at least for Bruker data - it should not be an issue any more. We played around with this in the past especially in connection with linear back prediction and it worked reasonably well. I must say your "ultimate solution" looks quite nice but for my taste the baseline still shows a bit of a curve. Have you ever tried the "baseopt" digital filter from Bruker? With this relatively new filtering method you don’t get smiles any more and the baseline is really flat. You also don’t see any vertical offset and there is one additional feature that really makes a difference: there is no need for any 1st order phase correction only 0th order phasing is required. We use baseopt on a regular basis for quite some time now, in particular for large series of metabonomics samples (urine, serum/plasma & CSF) where a high degree of reproducibility and a minimum of baseline and phasing artifacts are an absolute must. Baseopt allowed us to implement a fully automated processing workflow, so the only smile I usually see is on the face of the operator because he/she does not have to manually phase or baseline correct the spectra any more and studies with 2000-3000 samples (as it is typical in a clinical setting like ours) are easy feasible.

Carlos said...

Dear Axel,
Many thanks for your comprehensive comment. The group delay issue is certainly far from being a new one and if I remember well, first time I met with it was in 1996 whilst I was initiating my work on NMR data processing.
I admit that since Bruker added the length of the group delay into the acquisition parameters (acqus), things are now easier to handle as it’s no longer necessary to estimate this value, either using a look-up table or by means of some ad-hoc algorithm. However, in my opinion, this subject has a more philosophical background and whilst I agree that this is no longer a big problem and baselines (and phase) obtained with baseopt digital filter are now extremely good, IMHO, the fact that the FID shows this group delay is a sign of a fundamental problem which can yield to further signal processing difficulties. For example, Filter Diagonalization Method will not work directly on non time corrected FIDs. Of course, as you have pointed out with bLP, one can circumvent any of these problems, but if the FIDs were time corrected, one would not have to do any extra processing.
Thanks again for your comment!

Anonymous said...

To reiterate, Varian "neatly avoided the trap" -- and I assure you it was not "sheer luck"...

For years now, virtually all Varian spectra have had flat baselines, no "smiles" on the ends, no 1st order phase correction (in any dimension), and no vertical offset. Automatically. In the time-domain data. Due to "ddrtc".