Tehran University of Medical Sciences Frontiers in Biomedical Technologies 2345-5837 0 0 2026 06 10 1039 1039 Jasmine Christabel G Department of Computer Application, Noorul Islam Center for Higher Education, Kumaracoil, India A.C. Subhajini Department of Computer Application, Noorul Islam Center for Higher Education, Kumaracoil, India 2024 06 15 2025 04 06 Purpose: Diseases are unavoidable because of environmental factors, changes in diet, hereditary issues and many other factors; hence, it is important to detect diseases via various techniques in the healthcare sector to identify and diagnose the disease. Therefore, the proposed model focuses on employing advanced techniques for detecting heart disease, thyroid disease, and hepatitis, as these diseases have become common in recent years, along with the prediction of heart rate. Materials and Methods: The proposed work employs modified PCA (principal component analysis) for dimensionality reduction to extract appropriate features for the model by utilizing two learning rates (LR1 and L2). Furthermore, the modified back propagation (BP) method is used for effective classification of heart, thyroid, hepatitis, and heart rate prediction by incorporating adaptive Gaussian white noise (AWGN). In the proposed model, three different datasets are utilized: a heart disease dataset, a thyroid dataset, a hepatitis dataset for classification, and a heart rate prediction dataset for regression. Results: The accuracy, precision, recall, and F1 scores obtained by the proposed model for the heart disease dataset are 97.8%, 98%, 98%, and 98%, respectively. Similarly, 97.2%, 98%, 89%, and 93% for the thyroid dataset, respectively. Finally, the accuracy, precision, recall, and F1 score obtained by the proposed model for hepatitis are 95%, 98%, 88%, and 92%, respectively. Like the classification of diseases, heart rate prediction was also evaluated via different metrics, such as the RMSE, MSE, MAE, and R2. The MAE obtained by the proposed model for the heart rate prediction dataset is 0.112; likewise, the R2 obtained is 0.99, the MSE attained is 0.022, and the RMSE value obtained is 0.1488. Conclusion: The results of the proposed mechanism reflect its ability to detect different diseases effectively. This is due to the successful implementation of advanced AI approaches in the proposed framework. https://fbt.tums.ac.ir/index.php/fbt/article/view/1039 https://fbt.tums.ac.ir/index.php/fbt/article/download/1039/561