The advent of Internet-of-Things (IoT)-based telemedicine systems has ushered in a new era of technology facilitating early diagnosis and prevention for distant patients. This is particularly crucial for severe illnesses such as Alzheimer's disease, encompassing memory loss and cognitive dysfunction that significantly impairs daily life, necessitating immediate medical attention. The surge in data from intelligent systems, sourced from diverse locations, has heightened complexity and diminished diagnostic accuracy. In response, this study proposes an innovative distributed learning-based classification model, leveraging a deep convolutional neural network (CNN) classifier. This model proficiently manages clinical data images from disparate sources, ensuring disease classification with high accuracy. The research introduces a novel system designed for automated Alzheimer's disease detection and healthcare delivery. Comprising two subsystems, one dedicated to Alzheimer's diagnosis with an impressive 94.91% accuracy using CNN, and another for healthcare treatment, delivering excellent results. Notably, the system is adaptable to various diseases post-training. The study emphasizes the model's robust performance, achieving an outstanding 94.91% accuracy after 200 training epochs, with a loss of 0.1158, and a validation accuracy of 96.60% with a loss of 0.0922 at training without noise and loss: 0.2938 - Accuracy: 0.8713 - val_loss: 0.2387 - val_accuracy: 0.9069 at CNN with noise. Precision, recall, and F1 scores are comprehensively presented in a classification report, underscoring the system's effectiveness in categorizing Mild Demented and Non-Demented cases. While acknowledging room for further enhancements, this study introduces a promising avenue for telemedicine systems. It significantly impacts the early diagnosis and treatment of Alzheimer's disease and related medical conditions, thereby advancing the healthcare sector and improving patients' quality of life. The inclusion of these quantitative results enhances the abstract's appeal to readers, providing a clearer understanding of the study's outcomes.
Read full abstract