Hybrid deep learning framework for heart disease prediction using ECG signal images

With cardiovascular diseases accounting for all other causes of mortality worldwide, an increasing proportion of individuals are being treated for them. To identify the cardiac issue, medical practitioners have to examine electrocardiogram (ECG) data. Moreover, diagnosis is a process that healthcare professionals find simple vulnerable to mistakes. Modern deep learning systems have tackled the difficult choreography of independently identifying heart disease using ECGs. The present study is in line with the increasing frequency of cardiovascular disease, a major cause of death and sickness globally. From earlier and more precise detection of cardiac issues, improved patient outcomes and lower healthcare costs are conceivable. Although electrocardiogram (ECG) impulses are a diagnostic tool of great significance, their interpretation usually depends on professional analysis, which is prone to human variability and error. With an aim to lower computational complexity, this study proposes a hybrid deep learning framework for heart disease prediction utilizing artificial neural network models. This work aims to develop a deep learning approach using artificial neural networks (DLA-ANNs), automatically diagnoses heart disorders. The data have come from an ECG heartbeat classification kaggle dataset. The experimental results reveal that the ANN-based design provides superior accuracy than the state-of- the-art approaches. The outcomes reveal that the recommended strategy performs effectively, so it might be implemented in a medical setting. The proposed method surpasses other previously in use methods in several important respects: accuracy (93.6%), sensitivity (97.4%), adaptability (98.2%), performance (97.9%), and scalability (96.8%).