Automatic heart sound analysis with short-time Fourier transform and support vector machines

Author(s):  
Wen-Chung Kao ◽  
Chih-Chao Wei ◽  
Jen-Jui Liu ◽  
Pei-Yung Hsiao
2014 ◽  
Vol 6 (8) ◽  
pp. 2586-2591 ◽  
Author(s):  
Yuzhen Lu ◽  
Changwen Du ◽  
Changbing Yu ◽  
Jianmin Zhou

Fourier transform mid-infrared photoacoustic spectroscopy (FTIR-PAS) was employed to determine the contents of magnesium and potassium in rapeseeds.


Author(s):  
Achmad Rizal ◽  
Wahmisari Priharti ◽  
Sugondo Hadiyoso

Epilepsy is the most common form of neurological disease. The electroencephalogram (EEG) is the main tool in the observation of epilepsy. The detection and prediction of seizures in EEG signals require multi-domain analysis, one of which is the time domain combined with other approaches for feature extraction. In this study, a method for detecting seizures in epileptic EEG is proposed using analysis of the distribution of the signal spectrum in the time range t. The EEG signal which includes normal, inter-ictal and ictal is transformed into the time-frequency domain using the Short-Time Fourier Transform (STFT). Simulations were carried out on varying window length, overlap and FFT points to find the highest detection accuracy. The frequency distribution and first-order statistics were then calculated as feature vectors for the classification process. A support vector machine was employed to evaluate the proposed method. The simulation results showed the highest accuracy of 92.3% using 25-20-512 STFT and quadratic SVM. The proposed method in this study is expected to be a basis for the detection and prediction of seizures in long-term EEG recordings or real-time EEG monitoring of epilepsy patients.


2015 ◽  
Vol 15 (01) ◽  
pp. 1550009 ◽  
Author(s):  
KEHAN ZENG ◽  
ZHEN TAN ◽  
MINGCHUI DONG

A soft-computing method attenuating noise from heart sound (HS) signal for wearable e-healthcare device is proposed. The HS signal is decomposed by third-level wavelet packet transform (WPT). An automatic HS cycle detection algorithm is applied to find HS cycles in the (3, 0) leaf signal of WPT tree. Based on the quasi-cyclic feature of HS, short-time Fourier transform is implemented for cycles of each WPT tree leaf signal to decompose each cycle into time-frequency fragments which are called particles. Furthermore, the novel cuboid method is proposed to identify constituents of HS and noise from such generated particles. The particles representing HS are then retained and merged into noise-quasi-free WPT tree leaf signals. Eventually the inverse WPT (IWPT) is employed to build the noise-quasi-free HS signal. The method is assessed using mean square error (MSE) and compared with wavelet multi-threshold method (WMTM) and Tang's method. The experimental results show that the proposed method not only filters HS signal effectively but also well retains its pathological information.


2018 ◽  
Author(s):  
Bruno Soares da Silva ◽  
Gustavo Teodoro Laureano ◽  
Kleber Vieira Cardoso

A detecção acurada de indivíduos em ambientes fechados demanda dispositivos de alto custo, enquanto dispositivos de baixo custo, além da baixa acurácia, oferecem poucas informações sobre os eventos monitorados. As perturbações que podem afetar o sinal eletromagnético utilizado por interfaces de rede 802.11 tornam esse tipo de dispositivo um sensor de baixo custo, amplamente disponível e com acurácia satisfatória para várias aplicações. Neste trabalho, apresentamos o WiDMove, uma proposta para detecção da entrada e saída de pessoas em ambientes fechados utilizando medidas de qualidade do canal oferecidas pelo padrão IEEE 802.11n, conhecidas como Channel State Information (CSI). Nossa proposta é baseada em técnicas de processamento de sinal e de aprendizado de máquina, as quais nos permitem extrair e classificar assinaturas de eventos usando as medidas CSI. Em testes de laboratório com interfaces 802.11 convencionais, coletamos medidas CSI influenciadas por 8 pessoas distintas e extraímos as assinaturas de entrada e saída utilizando, dentre outras técnicas, Principal Component Analysis (PCA) e Short-Time Fourier Transform (STFT). Treinamos um classificador do tipo Support Vector Machine (SVM) e o validamos com validação cruzada, utilizando as técnicas K-Fold e Leave-One-Out. Os testes demonstraram que o WiDMove pode atingir a uma acurácia média superior a 85%. 


Sign in / Sign up

Export Citation Format

Share Document