Time–Frequency Transformations for Enhanced Biomedical Signal Classification with Convolutional Neural Networks

<b>Background:</b> Transforming one-dimensional (1D) biomedical signals into two-dimensional (2D) images enables the application of convolutional neural networks (CNNs) for classification tasks. In this study, we investigated the effectiveness of different 1D-to-2D transformation methods...

Full description

Saved in:
Bibliographic Details
Main Authors: Georgios Lekkas, Eleni Vrochidou, George A. Papakostas
Format: Article
Language:English
Published: MDPI AG 2025-01-01
Series:BioMedInformatics
Subjects:
Online Access:https://www.mdpi.com/2673-7426/5/1/7
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850204929526857728
author Georgios Lekkas
Eleni Vrochidou
George A. Papakostas
author_facet Georgios Lekkas
Eleni Vrochidou
George A. Papakostas
author_sort Georgios Lekkas
collection DOAJ
description <b>Background:</b> Transforming one-dimensional (1D) biomedical signals into two-dimensional (2D) images enables the application of convolutional neural networks (CNNs) for classification tasks. In this study, we investigated the effectiveness of different 1D-to-2D transformation methods to classify electrocardiogram (ECG) and electroencephalogram (EEG) signals. <b>Methods:</b> We select five transformation methods: Continuous Wavelet Transform (CWT), Fast Fourier Transform (FFT), Short-Time Fourier Transform (STFT), Signal Reshaping (SR), and Recurrence Plots (RPs). We used the MIT-BIH Arrhythmia Database for ECG signals and the Epilepsy EEG Dataset from the University of Bonn for EEG signals. After converting the signals from 1D to 2D, using the aforementioned methods, we employed two types of 2D CNNs: a minimal CNN and the LeNet-5 model. Our results indicate that RPs, CWT, and STFT are the methods to achieve the highest accuracy across both CNN architectures. <b>Results:</b> These top-performing methods achieved accuracies of 99%, 98%, and 95%, respectively, on the minimal 2D CNN and accuracies of 99%, 99%, and 99%, respectively, on the LeNet-5 model for the ECG signals. For the EEG signals, all three methods achieved accuracies of 100% on the minimal 2D CNN and accuracies of 100%, 99%, and 99% on the LeNet-5 2D CNN model, respectively. <b>Conclusions:</b> This superior performance is most likely related to the methods’ capacity to capture time–frequency information and nonlinear dynamics inherent in time-dependent signals such as ECGs and EEGs. These findings underline the significance of using appropriate transformation methods, suggesting that the incorporation of time–frequency analysis and nonlinear feature extraction in the transformation process improves the effectiveness of CNN-based classification for biological data.
format Article
id doaj-art-1e858ed4613441a2ac1abbfcc5a53ea6
institution OA Journals
issn 2673-7426
language English
publishDate 2025-01-01
publisher MDPI AG
record_format Article
series BioMedInformatics
spelling doaj-art-1e858ed4613441a2ac1abbfcc5a53ea62025-08-20T02:11:12ZengMDPI AGBioMedInformatics2673-74262025-01-0151710.3390/biomedinformatics5010007Time–Frequency Transformations for Enhanced Biomedical Signal Classification with Convolutional Neural NetworksGeorgios Lekkas0Eleni Vrochidou1George A. Papakostas2MLV Research Group, Department of Informatics, Democritus University of Thrace, 65404 Kavala, GreeceMLV Research Group, Department of Informatics, Democritus University of Thrace, 65404 Kavala, GreeceMLV Research Group, Department of Informatics, Democritus University of Thrace, 65404 Kavala, Greece<b>Background:</b> Transforming one-dimensional (1D) biomedical signals into two-dimensional (2D) images enables the application of convolutional neural networks (CNNs) for classification tasks. In this study, we investigated the effectiveness of different 1D-to-2D transformation methods to classify electrocardiogram (ECG) and electroencephalogram (EEG) signals. <b>Methods:</b> We select five transformation methods: Continuous Wavelet Transform (CWT), Fast Fourier Transform (FFT), Short-Time Fourier Transform (STFT), Signal Reshaping (SR), and Recurrence Plots (RPs). We used the MIT-BIH Arrhythmia Database for ECG signals and the Epilepsy EEG Dataset from the University of Bonn for EEG signals. After converting the signals from 1D to 2D, using the aforementioned methods, we employed two types of 2D CNNs: a minimal CNN and the LeNet-5 model. Our results indicate that RPs, CWT, and STFT are the methods to achieve the highest accuracy across both CNN architectures. <b>Results:</b> These top-performing methods achieved accuracies of 99%, 98%, and 95%, respectively, on the minimal 2D CNN and accuracies of 99%, 99%, and 99%, respectively, on the LeNet-5 model for the ECG signals. For the EEG signals, all three methods achieved accuracies of 100% on the minimal 2D CNN and accuracies of 100%, 99%, and 99% on the LeNet-5 2D CNN model, respectively. <b>Conclusions:</b> This superior performance is most likely related to the methods’ capacity to capture time–frequency information and nonlinear dynamics inherent in time-dependent signals such as ECGs and EEGs. These findings underline the significance of using appropriate transformation methods, suggesting that the incorporation of time–frequency analysis and nonlinear feature extraction in the transformation process improves the effectiveness of CNN-based classification for biological data.https://www.mdpi.com/2673-7426/5/1/7biomedical signal processingAI in healthcarebiomedical AIelectrocardiogram (ECG) classificationECGsdeep learning
spellingShingle Georgios Lekkas
Eleni Vrochidou
George A. Papakostas
Time–Frequency Transformations for Enhanced Biomedical Signal Classification with Convolutional Neural Networks
BioMedInformatics
biomedical signal processing
AI in healthcare
biomedical AI
electrocardiogram (ECG) classification
ECGs
deep learning
title Time–Frequency Transformations for Enhanced Biomedical Signal Classification with Convolutional Neural Networks
title_full Time–Frequency Transformations for Enhanced Biomedical Signal Classification with Convolutional Neural Networks
title_fullStr Time–Frequency Transformations for Enhanced Biomedical Signal Classification with Convolutional Neural Networks
title_full_unstemmed Time–Frequency Transformations for Enhanced Biomedical Signal Classification with Convolutional Neural Networks
title_short Time–Frequency Transformations for Enhanced Biomedical Signal Classification with Convolutional Neural Networks
title_sort time frequency transformations for enhanced biomedical signal classification with convolutional neural networks
topic biomedical signal processing
AI in healthcare
biomedical AI
electrocardiogram (ECG) classification
ECGs
deep learning
url https://www.mdpi.com/2673-7426/5/1/7
work_keys_str_mv AT georgioslekkas timefrequencytransformationsforenhancedbiomedicalsignalclassificationwithconvolutionalneuralnetworks
AT elenivrochidou timefrequencytransformationsforenhancedbiomedicalsignalclassificationwithconvolutionalneuralnetworks
AT georgeapapakostas timefrequencytransformationsforenhancedbiomedicalsignalclassificationwithconvolutionalneuralnetworks