Accuracy and Precision of Manual Baseline Determination

Abstract

Vibrational spectra often require baseline removal before further data analysis can be performed. Manual (i.e., user) baseline determination and removal is a common technique used to perform this operation. Currently, little data exists that details the accuracy and precision that can be expected with manual baseline removal techniques. This study addresses this current lack of data. One hundred spectra of varying signal-to-noise ratio (SNR), signal-to-baseline ratio (SBR), baseline slope, and spectral congestion were constructed and baselines were subtracted by 16 volunteers who were categorized as being either experienced or inexperienced in baseline determination. In total, 285 baseline determinations were performed. The general level of accuracy and precision that can be expected for manually determined baselines from spectra of varying SNR, SBR, baseline slope, and spectral congestion is established. Furthermore, the effects of user experience on the accuracy and precision of baseline determination is estimated. The interactions between the above factors in affecting the accuracy and precision of baseline determination is highlighted. Where possible, the functional relationships between accuracy, precision, and the given spectral characteristic are detailed. The results provide users of manual baseline determination useful guidelines in establishing limits of accuracy and precision when performing manual baseline determination, as well as highlighting conditions that confound the accuracy and precision of manual baseline determination.

SVD‐based method for intensity normalization, background correction and solvent subtraction in Raman spectroscopy exploiting the properties of water stretching vibrations

“…[23,24] These include manual baseline correction, [25] baseline fitting, [26] derivative methods [27] or data filtering (Fourier and wavelet transforms). [27,28] Nevertheless, methods based on factor analysis, principal component analysis or SVD seem to be rare.…”

“…For instance, polynomials shown in Fig. 2 were fitted to the series of points selected manually as user's best estimates [25] ; nevertheless more sophisticated methods can be employed for that purpose. [26] Afterwards, the baseline-corrected spectra Y i (ν) are constructed wileyonlinelibrary.com/journal/jrs Figure 2.…”

Low-Content Quantification in Powders Using Raman Spectroscopy: A Facile Chemometric Approach to Sub 0.1% Limits of Detection

“…[29][30][31][32][33][34] A chemometric approach (of which there are many) was preferred because this could be more easily implemented on conventional Raman systems. [35][36][37][38][39][40][41][42][43][44][45][46][47][48][49][50] Morphological weighted penalized least squares (MPLS) 49 was used for baseline correction of Raman spectra because of its inherent simplicity, combined with its flexibility, suitability for automation, and effectiveness at mitigating baseline artefacts. MPLS required neither a priori knowledge nor subjective user intervention, and was reasonably efficient computationally (~5 minutes for 8410 spectra).…”

Automatic Baseline Recognition for the Correction of Large Sets of Spectra Using Continuous Wavelet Transform and Iterative Fitting

“…9 Although the manual approach is still widely used, it is subjective, not perfectly reproducible, and time consuming. 10 Many different methodologies were developed to obtain an objective and reproducible baseline correction. A comprehensive description of most of them can be found in the work of Schulze et al 11 and a more recent one in the paper by Rowlands and Elliott.…”