AUTOMATED CARDIOVASCULAR DISEASES DIAGNOSIS WITH DEEP LEARNING MODEL

Cardiovascular disease (CVD) represents a significant global health burden that requires effective and timelydetection methods to ensure effective intervention. This project investigates the development of deep learningmodels for the early detection of cardiovascular disease using diverse and high-quality medical data. Throughcareful data processing and model selection, deep learning architectures, including MobileNet, convolutionalneural networks (CNNs), Xception and artificial neural networks (ANNs), are trained and optimized to accuratelyidentify patterns indicative of cardiovascular disease. Hyperparameter tuning and rigorous evaluation ensurerobust model performance in terms of evaluation of various metrics including precision, accuracy and recall. Inaddition, a user-friendly interface was developed to facilitate seamless interaction with the trained models,allowing users to enter important information and receive quick feedback on disease probability. This project aimsto advance the detection of cardiovascular disease by automating the process and reducing reliance on manualscreening, enabling early diagnosis and intervention, ultimately improving patient outcomes and reducing theburden of CVD on global health systems.Keywords:Cardiovascular Diseases, Electro Cardiogram, Machine Learning, Deep Learning, Convolutional NeuralNetworks, Artificial Neural Networks, Mean Absolute Error, Mean Squared Error, Artificial Intelligence, NaïveBayes, Arrhythmia, Congestive Heart Failure, Normal Sinus Rhythm, Continuous Wavelet Transform, SupportVector Machine, Graphics Processing Unit, Visual Geometry Group, Multi-Layer Perceptron, Fast ConditionalMutual Information, Matthews Correlation Coefficient, Data Flow Diagram, Recurrent Neural Networks,Abnormal Heartbeat, Myocardial Infarction, History of Myocardial Infarction, Normal Person, ReceiverOperating Characteristic Curve, Area Under Curve.