Hybrid ML Model for Early Detection of Neurological Disorders Using IoT-Based Health Monitoring Systems

Abstract

The increasing adoption of telemonitoring, wearable sensing, and Internet of Things (IoT) technologies has created new opportunities for the continuous monitoring and computational prediction of neurological disorders (NDs). In this study, a hybrid machine-learning framework is proposed, which integrates Adaptive Neuro-Fuzzy Inference System (ANFIS), Grey Wolf Optimization (GWO), and Particle Swarm Optimization (PSO) in an IoT- and fog-based health-monitoring architecture. A chaotic tent-mapping strategy is employed to initialize the optimization population and improve the diversity of candidate solutions. The proposed PSO-GWO method is used to optimize the parameters of the ANFIS model. The framework consists of data acquisition, preprocessing, feature extraction, classification, and alert-generation stages, with fog computing considered for processing health information closer to the data source. Five datasets from the UCI Machine Learning Repository are used for the experimental evaluation and the proposed approach is compared with GWO, PSO, Differential Evolution (DE), Genetic Algorithm (GA), Ant Colony Optimization (ACO) and conventional ANFIS. The experiments are conducted with a 70:30 training-testing split and multiple trials to assess model performance with error and classification metrics. The proposed method has been tested on the Parkinson's disease dataset and has been reported to have an accuracy of 92.5% under the reported experimental conditions. The results show that the hybrid optimization approach can enhance the classification performance of the ANFIS, but the proposed framework needs to be further tested with larger clinically representative datasets and real-time data from IoT/wearable devices before clinical use.