Evaluating the Efficiency of Deep Learning Models for Personalized Healthcare

The Internet of Medical Things (IoMT) has transformed healthcare through personalized care and the integration of medical equipment and data analytics. Deep learning models have shown promise in improving accuracy and efficiency in personalized healthcare. This study investigates the use of deep learning in IoMT, introducing the Butterfly-Optimized Deep Belief Network (BOBD) for disease diagnosis, therapy recommendation, and prognosis prediction. We use accuracy, precision, recall, F-measure, and RMSE metrics to compare BOBD's performance to that of K-NN, SVM, RF, and MLP. The findings demonstrate that the BOBD algorithm competes effectively with other classifiers. It outperforms K-NN and SVM in terms of accuracy, with a score of 0.90. BOBD outperforms K-NN and SVM in precision (0.86) and recall (0.84). The F-measure (0.85) demonstrates a balanced trade-off between precision and recall. While BOBD outperforms RF in accuracy and precision, it has a slightly higher RMSE (0.12) than RF (0.18), indicating a slightly higher prediction error. Our findings highlight the potential of deep belief networks optimized with the butterfly optimization method for personalized healthcare within IoMT. BOBD's satisfactory performance in classification accuracy and other assessment criteria demonstrates its utility in healthcare decision-making. Further research and optimization of BOBD hold the promise of improved outcomes and advancements in personalized healthcare.