Table 2.

ANOVAs’ results describing the significance of findings over the test configurations in comparison with our previous method⁴.

Variable	df	F	p	Sig.	${\eta }^2$
Accuracy
Algorithm for all phantoms (overall)	2	23.97	<0.001	*	0.48
Algorithm for perfusion phantoms	2	6.82	0.001	*	0.43
Algorithm for homogeneous phantoms	2	77.69	<0.001	*	0.89
Robustness
Local threshold	84	0.96	0.44		0.02
Global threshold	84	1.03	0.42		0.02
Reference configuration	12	10.38	0.007	*	0.21
Test configuration 1	12	7.86	0.016	*	0.14
Test configuration 2	12	5.95	0.031	*	0.15
Test configuration 3	12	8.12	0.015	*	0.07
Test configuration 4	12	10.09	0.008	*	0.23
Test configuration 5	12	11.14	0.006	*	0.12
Test configuration 6	12	7.05	0.021	*	0.16
Test configuration 7	12	5.5	0.037	*	0.13

The configurations are separated in accuracy and robustness tests. To ensure reproducibility of the p-values the following values are reported: Df = degrees of freedom, F = F-value, p = probability of the data given the null hypothesis, Sig. = p-values less than the traditional $\alpha$ <0.05 are marked with a “*”, ${\eta }^2$ = Generalized Eta-Squared measure of effect size.