Linear regression baseline correction

Linear Regression Baseline Fitting. This is a very simple approach to baseline correction in that it requires no effort to set up. In this method, a least squares regression line is fit to the responses in each spectral region selected for calibration. This line is then subtracted from the response values in the region before using the data to perform the calibration model calculations (Ref. 52). 

Different baseline correction methods vary with respect to the both the properties of the baseline component d and the means of determining the constant k. One of the simpler options, baseline ojfset correction, nses a flat-line baseline component (d = vector of Is), where k can be simply assigned to a single intensity of the spectrum x at a specific variable, or the mean of several intensities in the spectrum. More elaborate baseline correction schemes allow for more complex baseline components, such as linear, quadratic or user-defined functions. These schemes can also utilize different methods for determining k, such as least-squares regression. 

DOSY NMR experiments are popular for analysing diffusional properties of mixtures. Quality of the results is affected by experimental factors such as baseline drift, peak shift and phase shift. Huo et al proposed a series of pre-processing operations to reduce the experimental distortions, including baseline correction and reference deconvolution to remove frequency and phase shifts. The corrected data can be successfully analysed with a combination of multivariate curve resolution with non-linear least square regression. 

---