The following is an abstract for the selected article. A PDF download of the full text of this article is available here. Members may download full texts at no charge. Non-members may be charged a small fee for certain articles.
Accuracy and Precision of Manual Baseline Determination
Volume 58, Number 12 (Dec. 2004) Page 1488-1499
Jirasek, A.; Schulze, G.; Yu, M.M.L.; Blades, M.W.; Turner, R.F.B.
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.