Deep Learning-based Prediction of Cardiomegaly Disease from Thoracic X-ray Images using Convolutional Neural Networks

The abnormal expansion of the heart, or cardiomegaly, is a critical symptom of numerous cardiovascular disorders. The timely intervention and efficient patient care of cardiomegaly depend critically on the early and correct identification of the condition. In this article, we suggest a deep learning method for predicting cardiomegaly from thoracic X-ray images using convolutional neural networks (CNNs). We use pre-processing to extract pertinent features from the dataset as part of our methodology. Then, we create and train a CNN architecture that is especially suited for analysing thoracic X-rays. In order to improve the performance of the model, it is trained on a sizable dataset of annotated X-ray pictures using methods including transfer learning and fine-tuning. We carried out in-depth tests on a diversified dataset made up of X-ray pictures from various sources in order to assess the efficacy of our method. Our model achieved a notable accuracy of 85%, demonstrating its effectiveness in predicting the presence of cardiomegaly from thoracic X-ray images. The performance of our model is comparable to existing state-of-the-art methods in terms of sensitivity and specificity, further supporting its potential as a valuable tool for early detection of cardiovascular diseases. The findings of this research highlight the potential of deep learning and CNNs in the automated detection of cardiomegaly from thoracic X-ray images. The proposed approach demonstrates promising results and holds promise for future applications in clinical practice, offering a non-invasive and efficient method for early detection of cardiovascular diseases.