A New Approach to Diagnose Parkinson’s Disease Using a Structural Cooccurrence Matrix for a Similarity Analysis

Table 2

Comparison between the best results of this paper and the best results of the Pereira et al. approach [13].

Classif.	Database	Feature extractor	Accuracy (%)
Bayes	Meander	Pereira et al. approach [13]	59.20 ± 4.78
		Proposed approach	77.72 ± 4.27
	Spiral	Pereira et al. approach [13]	64.23 ± 7.11
		Proposed approach	82.01 ± 5.53
	M/S	Pereira et al. approach [13]	45.79 ± 4.15
		Proposed approach	71.39 ± 2.39
OPF	Meander	Pereira et al. approach [13]	57.54 ± 6.35
		Proposed approach	77.50 ± 2.75
	Spiral	Pereira et al. approach [13]	52.48 ± 5.32
		Proposed approach	75.71 ± 4.39
	M/S	Pereira et al. approach [13]	55.86 ± 3.63
		Proposed approach	67.21 ± 1.98
SVM	Meander	Pereira et al. approach [13]	66.72 ± 5.33
		Proposed approach	82.23 ± 3.02
	Spiral	Pereira et al. approach [13]	50.16 ± 1.71
		Proposed approach	85.54 ± 3.62
	M/S	Pereira et al. approach [13]	58.61 ± 2.84
		Proposed approach	74.13 ± 2.27