A Deep Pattern Learning based Model for Detection of Cardiovascular Diseases(CVD)

Given that cardiovascular diseases (CVD) persist as a significant global cause of death, the development of effective diagnostic techniques is of paramount importance. This article presents a novel method for detecting CVD using convolutional neural networks (CNN) and SqueezeNet, two cutting-edge de...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2024 4th International Conference on Pervasive Computing and Social Networking (ICPCSN) s. 191 - 196
Hlavní autoři: Srinivasulu, Balu, Reddy, P.V. Subba, Basha, P. Hussain
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 03.05.2024
Témata:
Adaptation models
Analytical models
Cardiovascular diseases
Computational modeling
Deep learning
Electrocardiogram (ECG)
Electrocardiography
Myocardium
Neural Network
Neural networks
Recurrent Neural Network (RNN)
SqueezeNet Convolutional Neural Network (CNN)
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Given that cardiovascular diseases (CVD) persist as a significant global cause of death, the development of effective diagnostic techniques is of paramount importance. This article presents a novel method for detecting CVD using convolutional neural networks (CNN) and SqueezeNet, two cutting-edge deep learning approaches. The proposed technique leverages ECG images to reliably classify medical images associated with cardiovascular problems. The foundation of feature extraction is SqueezeNet, a model renowned for its computational efficiency and lightweight construction. SqueezeNet's convolutional layers reduce the computing power required for model training and deployment and capture intricate patterns and textures indicative of CVD conditions, potentially revolutionizing the field of cardiovascular diagnostics. Furthermore, the pre-trained SqueezeNet model is honed using transfer learning methods on a collection of annotated cardiac images. This process allows the model to adapt its learned features to the specific characteristics of CVD images, thereby enhancing classification performance. Experimental evaluation on a benchmark dataset underscores the efficacy of the proposed approach in accurately identifying various types of cardiovascular abnormalities, including coronary artery disease, myocardial infarction, and heart valve defects. A comparati ve analysis against state-of-the-art methods demonstrates the superior performance and computational efficiency of the proposed S queezeNet-based model. It reinforces the confidence in its potential to transform the landscape of cardiovascular disease diagnosis. The findings of this study mark a significant stride in the development of computer-aided diagnosis systems for cardiovascular diseases, offering a promising tool for early detection and personalized treatment strategies.
DOI:10.1109/ICPCSN62568.2024.00040