Baseline-corrected spectral

Fig. 22.7 Quantitative evaluation of a baseline corrected spectral band.

Artifact removal and/or linearization. A common form of artifact removal is baseline correction of a spectrum or chromatogram. Common linearizations are the conversion of spectral transmittance into spectral absorbance and the multiplicative scatter correction for diffuse reflectance spectra. We must be very careful when attempting to remove artifacts. If we do not remove them correctly, we can actually introduce other artifacts that are worse than the ones we are trying to remove. But, for every artifact that we can correctly remove from the data, we make available additional degrees-of-freedom that the model can use to fit the relationship between the concentrations and the absorbances. This translates into greater precision and robustness of the calibration. Thus, if we can do it properly, it is always better to remove an artifact than to rely on the calibration to fit it. Similar reasoning applies to data linearization. 

After processing in the time domain, Fourier transformation, phasing and basic processing (calibration, peak picking, integration) ahs been performed, additional processing steps to improve spectral quality are at your disposal. This includes operations common to both ID and 2D spectra e.g. baseline correction in the frequency domain, as well as operations specific to these different type.s of data sets. 

In this section, we present the development of an automated protocol for prostate tissue histology [164] from infrared spectroscopic imaging data as an example of the techniques described (Fig. 8.11). The data is three dimensional with x-y—axes representing the image plane and the 2-axis representing the spectral dimension. After data acquisition, two important pre-processing steps, namely baseline correction and de-noising, are performed. Since the entire data set is derived from human tissue samples, the spectra have similar characteristics and, therefore, a manually chosen set of pre-defined wave number could be used as the reference points for baseline correction. It is... 

For physically uniform samples, simple techniques such as offset or polynomial baseline correction may suffice to minimize the physical/optical artifacts from the spectral data, but there are also a number of more advanced techniques available, which are quite effective. For example, it is common to convert the spectral data to a Savitzky-Golay33... 

Figure 8.2.15 One-dimensional spectra acquired with the four-coil probe. Each sample (250 mM in D2O) was loaded into the coil via the attached Teflon tubes 32 scans were acquired for each spectrum, with no delay between excitations of successive coils. Concurrent with the switch position being incremented, the spectral width was optimized for each compound 1 Hz line-broadening was applied before Fourier transformation and baseline correction. The spectral widths were (a) 600 Hz (galactose) (b) 1400 Hz (adenosine triphosphate) (c) 2000 Hz (chloroquine) (d) 500 Hz (fructose). 2048 complex data points were acquired for each spectrum, giving data acquisition times of approximately 1.7, 0.7, 0.5 and 2.0 s, respectively. The delay between successive 90 degree excitations was 4.9 s for each sample. Reprinted with permission From Li, Y., Walters, A., Malaway, P., Sweedler, J. V. and Webb, A. G., Anal. Chem.,l, 4815-4820 (1999). Copyright (1999) American Chemical Society...

Phase 2 - data preprocessing. There are many ways to process spectral data prior to multivariate image reconstruction and there is no ideal method that can be generally applied to all types of tissue. It is usual practice to correct the baseline to account for nonspecific matrix absorptions and scattering induced by the physical or bulk properties of the dehydrated tissue. One possible procedure is to fit a polynomial function to a preselected set of minima points and zero the baseline to these minima points. However, this type of fit can introduce artifacts because baseline variation can be so extreme that one set of baseline points may not account for all types of baseline variation. A more acceptable way to correct spectral baselines is to use the derivatives of the spectra. This can only be achieved if the S/N of the individual spectra is high and if an appropriate smoothing factor is introduced to reduce noise in the derivatized spectra. Derivatives serve two purposes they minimize broad... 

Fig. 4.1.7. Baseline-corrected and normalized FTIR spectra illustrating spectral types G, GS 1-GS 4, and HGS. G Norway spruce (Picea abies)-, GS I (Gnetum venosum) GS 2 lauan (Shorea polysperma), GS 3 dabema (Piptadeniastrum africanum) GS 4 birch (Betula sp.) HGS a bamboo (Bambusa sp.) b barley (Hordeum vulgare). (Instrument FTS 40 Bio-Rad, Digilab, 4cm 1 resolution, 32 scans, KBr pellet technique). According to Faix 1991...

Use baseline corrections on each integral cut if spectral tilt is noticeable. Extensive zero filling (see step 2) is useful, because it provides more data points to define line shapes better (Section 2-5b). Slight line broadening produces smoother spectral and integral lines with less baseline noise. 

FIGURE 21 Dissolution profiles for buffered aspirin tablets obtained with the fiber-optic dissolution system (260-350 nm) and manual sampling with HPLC analysis. The dissolution was performed with USP apparatus 2 at 75 revolutions per minute. A second-derivative baseline correction was performed on the fiber-optic raw spectral data to correct for scattering due to the turbid solution. 

---