Skip to main content

Table 4 Comparison with related works in 5 years using the same heart sound dataset

From: Diagnosis of cardiac abnormalities based on phonocardiogram using a novel fuzzy matching feature extraction method

References

Feature extraction

Classifier

Accuracy (%)

Sensitivity (%)

Specificity (%)

[8]

DWT + MFCCs

SVM

97.9

98.2

99.4

[8]

DWT

SVM

92.3

92.3

98.4

[8]

MFCCs

SVM

91.6

87.3

96.6

[5]

–

Deep wavenet

98.2

97.0

92.5

[31]

PCA

Random forest

94.8

94.7

98.7

[31]

PCA

KNN

91.6

91.5

97.9

[37]

–

CNN

98.7

98.7

99.6

This study

FMFE

SVM

94.9

95.7

98.8

 

FMFE

KNN

90.0

90.5

97.5

 

FMFE

Random forest

96.5

97.0

99.1

 

FMFE

MLP

93.1

93.7

98.3

 

FMFE + MFCCs

SVM

99.0

99.4

99.7

 

FMFE + MFCCs

KNN

98.4

98.5

99.6

 

FMFE + MFCCs

Random forest

98.2

98.5

99.6

 

FMFE + MFCCs

MLP

96.9

97.0

99.2

  1. –Denotes that there is no feature extraction method was used