IoT-Based Diagnosis and Recommendation System for Chronic Diseases Using Patient Health Records

The proliferation of IoT applications, notably in the sphere of health care, has led to discussion of patient health records utilising data gathered from IoT-connected devices. Biological data from patients' medical records is mined for health analysis and diagnosis. Certain types of illness, known as chronic illnesses, are completely quiet yet devastating if left untreated. Recent years have seen an uptick in interest from academics in the utilization of patient health information data for the pre-emptive identification of chronic diseases. On the other side, healthcare and medical assistance have benefited enormously from the implementation of recommender systems that use machine learning techniques. Using an Internet of Things device, this research implemented a medical recommendation system to aid in the early diagnosis and management of chronic conditions. The current technique made use of the dataset of digitised patient health records that is housed in the Physio Net data repository. The current dataset contains patient health records that have been documented in accordance with the diseases that have been identified and the doctor's diagnosis. The recommended technique uses K-nearest neighbour classification to identify the kind of ailment before using collaborative filtering to select the optimal course of treatment for patients. The outcomes of using the suggested methodology show that this technique, which is based on using patient symptom similarity, produces better results than previous methods and has a high precision in diagnosing and forecasting chronic illnesses. After determining the kind of disease using the closest neighbour classification method, the collaborative filtering strategy is utilised to select the optimal course of therapy for patients. The outcomes of using the suggested methodology show that this technique, which is based on using patient symptom similarity, produces better results than previous methods and achieves better accuracy in diagnosing and forecasting chronic illnesses.