Heart Sounds Classification Using Short-Time Fourier Transform and Gray Level Difference Method

Rizal, Achmad; Handzah, Vivi Aliyah Putri; Kusuma, Purba Daru

doi:10.18280/isi.270302

Cited by 7 publications

(8 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…From the process above, we would get an image with a size of m x n with a range of values from 0-255, which was equivalent to a grey-scale image with a depth of 8 bits [15]. Converted images are shown in Figure 4 and Figure 5.…”

Section: B Signal Conversionmentioning

confidence: 99%

Grey Level Differences Matrix for Alcoholic EEG Signal Classification

Sri Aprillia,

Rizal,

Geraldy Fauzi

2024

JOIV : Int. J. Inform. Visualization

View full text Add to dashboard Cite

Electroencephalogram (EEG) signals can provide information on abnormalities in a person's brain and characterize brain activity. Brain injury or diseases can manifest as brain disorders. Trauma or the use of specific chemicals or medications, such as alcohol, can result in brain damage. Previous research has demonstrated variations in the patterns of EEG signals between alcohol-using and non-drinking people. Various techniques, including wavelet and entropy, have been developed to detect alcoholic EEG using event-related potential (ERP) testing. This work proposes a feature extraction technique based on texture analysis for the classification of alcohol EEG signals because ERP-measured EEG often involves many channels. An NxM image is thought to be equivalent to an EEG signal with N channels and a recording duration of M samples. The NxM matrix is formed by channelizing the N-channel EEG signal in this investigation. Normalization is then used to get a matrix value of 0-255 or an 8-bit image in the following step. Five features are measured in four directions, and the Grey Level Difference Matrix (GLDM) approach is utilized for feature extraction. Using five grey-level difference matrix (GLDM) features and linear discriminant analysis as a classifier, the maximum accuracy was achieved at 73.3%. Image processing can still be used to increase accuracy even though the final product is less accurate than the earlier technique. The suggested approach can still be adjusted to work with biomedical signals or image processing techniques like the Grey Level Co-occurrence Matrix (GLCM).

show abstract

Section: B Signal Conversionmentioning

confidence: 99%

Grey Level Differences Matrix for Alcoholic EEG Signal Classification

Sri Aprillia,

Rizal,

Geraldy Fauzi

2024

JOIV : Int. J. Inform. Visualization

View full text Add to dashboard Cite

show abstract

“…To establish the credibility of the classification model, we undertook comparative experiments against the latest developments in the field and conducted ablation studies. About the work by the Rizal research team (Rizal et al, 2022) Frontiers in Physiology frontiersin.org…”

Section: Analysis Of Comparative Ablation Experimentsmentioning

confidence: 99%

Assistive diagnostic technology for congenital heart disease based on fusion features and deep learning

Wang,

Yang,

Qian

et al. 2023

Front. Physiol.

View full text Add to dashboard Cite

Introduction: Congenital heart disease (CHD) is a cardiovascular disorder caused by structural defects in the heart. Early screening holds significant importance for the effective treatment of this condition. Heart sound analysis is commonly employed to assist in the diagnosis of CHD. However, there is currently a lack of an efficient automated model for heart sound classification, which could potentially replace the manual process of auscultation.Methods: This study introduces an innovative and efficient screening and classification model, combining a locally concatenated fusion approach with a convolutional neural network based on coordinate attention (LCACNN). In this model, Mel-frequency spectral coefficients (MFSC) and envelope features are locally fused and employed as input to the LCACNN network. This model automatically analyzes feature map energy information, eliminating the need for denoising processes.Discussion: The proposed classification model in this study demonstrates a robust capability for identifying congenital heart disease, potentially substituting manual auscultation to facilitate the detection of patients in remote areas.Results: This study introduces an innovative and efficient screening and classification model, combining a locally concatenated fusion approach with a convolutional neural network based on coordinate attention (LCACNN). In this model, Mel-frequency spectral coefficients (MFSC) and envelope features are locally fused and employed as input to the LCACNN network. This model automatically analyzes feature map energy information, eliminating the need for denoising processes. To assess the performance of the classification model, comparative ablation experiments were conducted, achieving classification accuracies of 91.78% and 94.79% on the PhysioNet and HS databases, respectively. These results significantly outperformed alternative classification models.

show abstract

“…The algorithm for further learning is trained on one of the sets and then tested on the other. Sensitivity, specificity, and total accuracy are used to evaluate performance [33].…”

Section: Datasetsmentioning

confidence: 99%

Predicting Cancer Survival Using Multilayer Perceptron and High-Dimensional SVM Kernel Space

Kumar¹,

Khatri²,

Mohammadian³

2022

ISI

View full text Add to dashboard Cite

Predicting disease prediction and prognosis have become very easy task with Machine learning models for an inextricable aspect of cancer research aiming at enhancing patient therapy and management. The primary goal of proposed research work is to use Support vector machine, machine learning models and for dealing very accurately for predicting survival time for breast cancer based on clinical data. The study has proposed a solution to the problem in respect to various tumor related characteristic by integrating from dataset about tumour stage, size of tumor, and age at which the diagnosis start is an important major component for utilising for predicting survival time. Haberman's Survival Data Set is dataset describing those subjects who had been provided treatment for breast cancer. The sample taken for research are taken from study which was conducted at University of Chicago's Billings Hospital taking case who were survived after surgery for breast cancer. SVM applied on data set by different options of kernel RBF and linear as well as soft computing techniques are applied to predict the survival rate of patient from dataset. Apart from data standardisation and categorization, the machine learning approaches used in this research work to demonstrate features in terms of predicting how long they survived. Model performance is analysed on breast cancer data is justified by accuracy, support and f1 score. A workflow based on Python-platform has been utilised to support the suggested technique.

show abstract

Heart Sounds Classification Using Short-Time Fourier Transform and Gray Level Difference Method

Cited by 7 publications

References 19 publications

Grey Level Differences Matrix for Alcoholic EEG Signal Classification

Grey Level Differences Matrix for Alcoholic EEG Signal Classification

Assistive diagnostic technology for congenital heart disease based on fusion features and deep learning

Predicting Cancer Survival Using Multilayer Perceptron and High-Dimensional SVM Kernel Space

Contact Info

Product

Resources

About