Pathological Brain Detection Using Weiner Filtering, 2D-Discrete Wavelet Transform, Probabilistic PCA, and Random Subspace Ensemble Classifier

Table 4

Classification comparison with DS-90.

Existing methods	Success cases	Sensitivity (%)	Specificity (%)	Precision (%)	Accuracy (%)
DWT + PCA + BPNN [25]	388	88.00	56.00	97.14	86.22
DWT + PCA + RBF-NN [25]	411	92.47	72.00	98.25	91.33
DWT + PCA + PSO-KSVM [25]	440	98.12	92.00	99.52	97.78
WE + BPNN [26]	390	88.47	56.00	97.16	86.67
WE + KSVM [27]	413	93.18	68.00	98.02	91.78
DWT + PCA + GA-KSVM [28]	439	97.88	92.00	99.52	97.56
WE + PSO-KSVM [29]	437	97.65	88.00	99.28	97.11
WE + BBO-KSVM [29]	440	98.12	92.00	99.52	97.78
WE + QPSO-KSVM [30]	442	98.59	92.00	99.52	98.22
WFRFT + PCA + GEPSVM [31]	446	99.53	92.00	99.53	99.11
HMI + SEPSVM [15]	445	99.06	96.00		98.89
HMI + TSVM [15]	445	99.29	92.00		98.89
Proposed
2D- DWT + PPCA + RSE (proposed)	450	100.00	100.00	100.00	100.00