Investigation of glottal flow parameters for voice pathology detection on SVD and MEEI databases

Investigation and Evaluation of Glottal Flow Waveform for Voice Pathology Detection

“…However, the proposed method achieved 100% accuracy with the combined feature set extracted from glottal flow waveforms in the same subset of the MEEI corpus. In addition, the method [4] that only used glottal flow parameters achieved the lowest accuracy compared to other methods. This result also verifies the finding, noticed above, that conventional glottal source features can be effective complementary features, but not sufficient to represent an outstanding individual feature set.…”

“…MFCC was used as input parameter for DBSCAN-SVM classification, and 98.63% accuracy was obtained [19]. Ezzine et al achieved 93.66% accuracy using glottal flow parameters with ANN and SVM classification [4]. However, the proposed method achieved 100% accuracy with the combined feature set extracted from glottal flow waveforms in the same subset of the MEEI corpus.…”

“…Features used Classifier Database Best accuracy(%) [10] Autocorrelation and entropy SVM MEEI, SVD and AVPD 99.69, 92.79 and 99.79 [21] Glottal flow parameters SVM MEEI and SVD 99.27 and 93.66 [19] MFCC DBSCAN-SVM MEEI 98.63 [8] Jitter, Shimmer, HNR, TNI and NFHE KNN MEEI 96.1 [11] PPE SVM Private 91.4 [12] Largest Lyapunov exponent SVM MEEI 88.89 [22] openSMILE features and Glottal parameters SVM and CNN UA-speech and TORGO 87.93 and 76.66 [16] MFCC SVM SVD 86 [20] MFCC-QCP and Glottal source features SVM HUPA and SVD 78.37 [17] MLSF GMM MEEI 77.9 [9] HNR, NNE and GNE KNN UAM 66.57 which can represent the influence of voice diseases on the mechanism of the vocal folds.…”

“…The Gaussian mixture model (GMM), hidden Markov model (HMM), support vector machine (SVM) and random forest (RF) methods have been widely used due to their outstanding performances, especially SVM [3]. With the development of neural network technology, methods such as artificial neural networks (ANNs) [4], deep neural networks (DNNs) [5] and convolutional neural networks (CNNs) [6] have also been applied to the study of pathological voice. More discussion and description of classifier algorithms used for voice pathology detection and clasification can be obtained in recent review [3].…”

“…Hence, glottal source features that are extracted from the glottal flow waveform by using glottal inverse filtering are of great research value. In [21], the performance of glottal flow features, especially for distinguishing patients with benign and malignant tumors from healthy people, was investigated in the MEEI and SVD databases. ANN and SVM were used to classify normal and pathological voices, and the best classification rates were 99.27% and 93.66% for the two databases, respectively.…”