Table 3.

Summary of surface fitting algorithms available in NeuroMeasure.

Algorithm	Summary
Piecewise cubic spline	3rd order 2D polynomial fit to dataset in patches of 3 data points each, fit in a piecewise fashion directly without least squares. Edges between patches are smoothed by splines.
Local linear Weighted scatter Smoothing; a.k.a Lowess	1st order 2D polynomial fit to dataset via least squares regression combined with a ratio that splits the dataset into local parts and the polynomial is fit in a piecewise fashion. Ratio controls the degree of smoothing. Here, the ratio is fixed to 0.25 (0 = least smooth, 1 = most smooth)
Biharmonic (v4)	The same algorithm used in MATLAB's griddata function. See the comprehensive explanation on the Mathworks website griddata documentation: https://www.mathworks.com/help/matlab/ref/griddata