Optimizing Telemedicine Patient Care with Machine Learning for Disease Progression and Treatment Planning
K. Kokulavani, Kishore Varma Manthena, B Sakthisaravanan, Tammineni Sreelatha, J. Visumathi, K S Guruprakash, S Senthil Murugan · Engineering Technology & Applied Science Research · 2025
Telemedicine has revolutionized healthcare delivery, providing patients with easy access to medical services while reducing the need for in-person consultations. This study examines enhancing telemedicine patient care using Machine Learning (ML) techniques, particularly K-Nearest Neighbors (KNN) and Gradient Boosting Machines (GBM). The KNN method enables efficient patient classification by analyzing commonalities in health data, supporting personalized therapy recommendations customized to unique patient profiles. The simplicity and interpretability of KNN provide an attractive option for real-time telemedicine applications that do not require explicit training. GBM is used to overcome the limitations of traditional models such as Logistic Regression (LR) and Random Forest by enhancing prediction accuracy using an ensemble method that integrates many weak learners. The proposed GBM model uses 150 decision trees as base classifiers and aggregates their outputs for the final prediction. The proposed system uses the MIMIC-III dataset for performance evaluation and a five-fold stratified cross-validation. The GBM model achieved 97.56% accuracy and outperformed KNN (91.23%), LR (88.27%), and RF (94.35%). This study highlights the significant potential of ML in enhancing telemedicine procedures, facilitating better patient outcomes, and more efficient healthcare delivery.