Evaluation of Heart Disease Risk Using Deep Learning Technique with Image Enhancement

This study emphasizes the significance of the heart in the human body.Numerous serious vascular conditions exist in the heart and the blood.The dataset, study goals, methodology, approach, and efficient algorithms for identifying and classifying electrocardiogram (ECG) data are all covered in this paper.Picture scaling, grayscale conversion, and training/testing dataset segmentation are part of the cardiovascular ECG image-processing process.To assess ECG images, researchers used a convolutional neural network.Iterations in model training increase the accuracy.We examined generalization and model recall using an additional dataset.The accuracy, F1 score, confusion matrices, and ECG pattern identification must all be evaluated to diagnose heart disease.We used a Cardiovascular ECG Image Collection that was openly accessible.The system was constructed in Python using Matplotlib, NumPy, and Keras.The GPU-based machine learning platform was Google Colab.Photos were analyzed, categorized, and processed using MobileNet-V2.With a remarkable accuracy rate of 99.3 %, the developed model offers a viable basis for further hyperparameter investigation.Additionally, to demonstrate the efficacy of the selected method, we present a graphic depiction of ECG data.Overall, this study combines advanced machine learning algorithms, strict assessment criteria, and ECG image analysis to enhance the diagnosis of heart-related disorders.