Table 3.

The ten most discriminating parameters according to the Fisher (F-) coefficient and corresponding Wilcoxon test p-values.

Mesencephalon
Most Discriminative Texture Parameters on Patients	p-values (patients)	p-values (controls)	Most Discriminative Texture Parameters on Controls	p-values (controls)	p-values (patients)
Teta4	<0.001*	0.105	S(4,-4)DifVarnc	0.002*	0.519
Teta3	<0.001*	0.020*	S(5,-5DifVarnc	0.004*	0.465
S(5,5)Entropy	<0.001*	0.105	Teta3	0.020*	<0.001*
S(5,5)AngScMom	<0.001*	0.020*	S(5,-5)DifEntrp	0.004*	0.413
S(4,4)AngScMom	<0.001*	0.432	Teta2	0.014*	<0.001*
S(4,4)Entropy	<0.001*	0.322	S(4,-4)DifEnrtp	0.002*	0.372
Teta2	<0.001*	0.014*	S(3,-3)DifVarnc	0.002*	0.160
S(1,-1)DifVarnc	<0.001*	0.375	WavEnLL_s-2	0.027*	0,472
S(1,1)DifVarnc	<0.001*	0.375	S(5,-5)Contrast	0.002*	0.833
S(3,3)AngScMom	<0.001*	0.492	S(4,-4)Contrast	0.002*	0.432

The ten most discriminating parameters according to the Fisher (F-) coefficient and their corresponding Wilcoxon test p-values are calculated between hemispheres in the area of mesencephalon for both patients and controls.

Teta, vectors of autoregressive model; AngScMom, angular second moment; SumAverg, sum average; DifEntrp, difference entropy; DiffVarnc, difference variance; WavEnLL, energy of wavelet coefficients in subband LL.

* p-values of under 0.05 are considered statistically significant