A generative model for Deep Fake Augmentation of phonocardiogram and electrocardiogram signals using LSGAN and Cycle GAN

• Autorzy: Rayavarapu Swarajya Madhuri , Prasanthi Tammineni Shanmukha , Kumar Gottapu Santosh , Rao Gottapu Sasibhushana , Prashanti Gottapu

Identyfikatory

DOI

10.35784/iapgos.3783

Warianty tytułu

PL

Generatywny model z Deep Fake Augumentation dla sygnałów z fonokardiogramu oraz elektrokardiogramu w strukturach LSGAN oraz Cycle GAN

• Języki publikacji: EN

Abstrakty

EN

In order to diagnose a range of cardiac conditions, it is important to conduct an accurate evaluation of eitherphonocardiogram (PCG)and electrocardiogram (ECG) data. Artificial intelligence and machine learning-based computer-assisted diagnostics are becoming increasingly commonplace in modern medicine, assisting clinicians in making life-or-death decisions. The requirement for an enormous amount of informationfor training to establish the framework for a deeplearning-based technique is an empirical challenge in the field of medicine. This increases the riskof personal information being misused. As a direct result of this issue, there has been an explosion in the study of methods for creating synthetic patient data. Researchers have attempted to generate synthetic ECG or PCG readings. To balance the dataset, ECG data were first created on the MIT-BIH arrhythmia database using LS GAN and Cycle GAN. Next, using VGGNet, studies were conducted to classify arrhythmias for the synthesized ECG signals. The synthesized signals performed well and resembled the original signal and the obtained precision of 91.20%, recall of 89.52% and an F1 scoreof 90.35%.

PL

W celu zdiagnozowania szeregu chorób serca, istotne jest przeprowadzenie dokładnej oceny danych z fonokardiogramu (PCG)i elektrokardiogram (EKG). Sztuczna inteligencja i diagnostyka wspomagana komputerowo, oparta na uczeniu maszynowym stają sięcoraz bardziej powszechne we współczesnej medycynie, pomagając klinicystom w podejmowaniu krytycznych decyzji. Z kolei, Wymóg ogromnej ilości informacjido trenowania, w celu ustalenia platformy (ang. framework) techniki, opartej na głębokim uczeniu stanowi empiryczne wyzwanie w obszarze medycyny. Zwiększa to ryzyko niewłaściwego wykorzystania danych osobowych. Bezpośrednim skutkiem tego problemu był gwałtowny rozwój badań nad metodami tworzenia syntetycznych danych pacjentów. Badacze podjęli próbę wygenerowania syntetycznych odczytów diagramów EKG lub PCG. Stąd, w celu zrównoważenia zbioru danych, w pierwszej kolejności utworzono dane EKG w bazie danych arytmii MIT-BIH przy użyciu struktur sieci generatywnych LSGAN i CycleGAN. Następnie, wykorzystując strukturę sieci VGGNet, przeprowadzono badania, mające na celu klasyfikację arytmii na potrzeby syntetyzowanych sygnałów EKG. Dla wygenerowanych sygnałów, przypominających sygnał oryginalny uzyskano dobre rezultaty. Należy podkreślić,że uzyskana dokładność wynosiła 91,20%, powtarzalność 89,52% i wynik F1 –odpowiednio 90,35%.

Słowa kluczowe

EN

arrhythmia; auscultation; electrocardiogram; phonocardiogram; generative networks

PL

arytmia; osłuchiwanie; elektrokardiogram; fonokardiogram; sieci generatywne

• Wydawca: Wydawnictwo Politechniki Lubelskiej
• Czasopismo: Informatyka, Automatyka, Pomiary w Gospodarce i Ochronie Środowiska
• Rocznik: 2023
• Tom: T. 13, nr 4
• Strony: 34--38
• Opis fizyczny: Bibliogr. 27 poz., rys., tab., wykr.

Twórcy

autor

Rayavarapu Swarajya Madhuri

• Andhra University, Department of Electronics and Communication Engineering, Visakhapatnam,India

• https://orcid.org/0009-0007-7559-2142

autor

Prasanthi Tammineni Shanmukha

• Andhra University, Department of Electronics and Communication Engineering, Visakhapatnam,India

• https://orcid.org/0009-0000-5352-2265

autor

Kumar Gottapu Santosh

• Gayatri Vidya Parishad College of Engineering, Department of Civil Engineering,Visakhapatnam, India

• https://orcid.org/0000-0002-1452-9752

autor

Rao Gottapu Sasibhushana

• Andhra University, Department of Electronics and Communication Engineering, Visakhapatnam,India

• https://orcid.org/0000-0001-6346-8274

autor

Prashanti Gottapu

• Avanthi Institute of Pharmaceutical Sciences, Department ofPharmaceutical Technology, Vizianagaram, India

• https://orcid.org/0009-0002-7231-0377

Bibliografia

• [1] Ahmed N., Zhu Y.: Early Detection of Atrial Fibrillation Based on ECG Signals. Bioengineering 7(1), 2020, 16 [http://doi.org/10.3390/bioengineering7010016].
• [2] Akkaradamrongrat S. et al.: Text generation for imbalanced text classification. 16th International Joint Conference on Computer Science and Software Engineering (JCSSE), 2019, 181–186.
• [3] Aziz S. et al.: Phonocardiogram Signal Processing for Automatic Diagnosis of Congenital Heart Disorders through Fusion of Temporal and Cepstral Features. Sensors 20(13), 2020, 3790 [http://doi.org/10.3390/s20133790].
• [4] Bentley P. et al.: Classifying Heart Sounds Challenge. 2011 [http://www.peterjbentley.com/heartchallenge/index.html]
• [5] Bouril D. et al.: Automated classification of normal and abnormal heart sounds using support vector machines. Computing in Cardiology Conference – CinC, Vancouver 2016, 549–552.
• [6] Cayce G. I. et al.: Improved Neural Network Arrhythmia Classification Through Integrated Data Augmentation. IEEE MetroCon, Hurst 2022, 1–3.
• [7] England J. R., Cheng P. M.: Artificial intelligence for medical image analysis: a guide for authors and reviewers. American journal of roentgenology 212(3), 2019, 513–519.
• [8] Garcea F. et al.: Data augmentation for medical imaging: A systematic literature review. Computers in Biology and Medicine 152, 2023, 106391 [http://doi.org/10.1016/j.compbiomed.2022.106391].
• [9] Goldberger A. L. et al.: PhysioBank, PhysioToolkit and PhysioNet: components of a new research resource for complex physiologic signals. Circulation 101, 2000, 215–220.
• [10] Goodfellow I. et al.: Generative adversarial networks. Communications of the ACM, 63(11), 2020, 139–144 [http://doi.org/10.1145/3422622].
• [11] Guo G. et al.: Multimodal Emotion Recognition Using CNN-SVM with Data Augmentation. IEEE International Conference on Bioinformatics and Biomedicine, Las Vegas 2022, 3008–3014.
• [12] Houssein E. H.: ECG signals classification: a review. International Journal of Intelligent Engineering Informatics 5(4), 2017, 376–396.
• [13] Judge R., Mangrulkar R.: Heart Sound and Murmur Library. [http://open.umich.edu/education/med/resources/heart-sound-murmur-library/2015].
• [14] Khalifa Y et al.: A review of Hidden Markov models and Recurrent Neural Networks for event detection and localization in biomedical signals. Information Fusion 69, 2021, 52–72.
• [15] Li H. et al.: Dual-input neural network integrating feature extraction and deep learning for coronary artery disease detection using electrocardiogram and phonocardiogram. IEEE Access 7, 2019, 146457–146469.
• [16] Li J., Ke L., Du Q., Ding X., Chen X.: Research on the Classification of ECG and PCG Signals Based on BiLSTM-GoogLeNet-DS. Applied Sciences 12(22), 2022, 11762.
• [17] Liu C. et al.: An open access database for the evaluation of heart sound algorithms. Physiological Measurement 37(12), 2016, 2181.
• [18] Mao X. et al.: Least Squares Generative Adversarial Networks. arXiv, 2017 [http://arxiv.org/abs/1611.04076].
• [19] Nedoma J. et al.: Comparison of BCG, PCG and ECG signals in application of heart rate monitoring of the human body. 40th International Conference on Telecommunications and Signal Processing – TSP, 2017, 420–424.
• [20] Rahman, M. M. et al.: A Systematic Survey of Data Augmentation of ECG Signals for AI Applications. Sensors, 23(11), 2023, 5237 [http://doi.org/10.3390/s23115237].
• [21] Simonyan K., Zisserman A.: Very Deep Convolutional Networks for Large-Scale Image Recognition. arXiv, 2015 [http://arxiv.org/abs/1409.1556].
• [22] Skandarani Y. et al.: GANs for medical image synthesis: An empirical study. Journal of Imaging 9(3), 2023 [http://doi.org/10.3390/jimaging9030069].
• [23] Sreeniwas Kumar A., Nakul S.: Cardiovascular Disease in India: A 360 Degree Overview. Medical Journal Armed Forces India 76(1), 2020, 1–3 [http://doi.org/10.1016/j.mjafi.2019.12.005].
• [24] Wang T. C. et al.: High-resolution image synthesis and semantic manipulation with conditional gans. IEEE Conference on computer vision and pattern recognition. Salt Lake City, 2018, 8798–8807.
• [25] Wu J. L. et al.: A prediction model of stock market trading actions using generative adversarial network and piecewise linear representation approaches. Soft Comput 27, 2023, 8209–8222
• [http://doi.org/10.1007/s00500-022-07716-2].
• [26] Xiong P. et al.: Deep Learning for Detecting and Locating Myocardial Infarction by Electrocardiogram: A Literature Review. Frontiers in Cardiovascular Medicine 9, 2022 [http://doi.org/10.3389/fcvm.2022.860032].
• [27] Zhu J. Y. et al.: Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks. arXiv [http://arxiv.org/abs/1703.10593].

Uwagi

Opracowanie rekordu ze środków MNiSW, umowa nr SONP/SP/546092/2022 w ramach programu "Społeczna odpowiedzialność nauki" - moduł: Popularyzacja nauki i promocja sportu (2024).