A Hybrid IoT-Enabled Transfer Learning Model for Accurate Diagnosis of Coronary Heart Disease

Abstract

Coronary heart disease (CHD) is the top cause of death in the world, which shows how important it is to get a correct evaluation as soon as possible. The Internet of Things (IoT) opens up new ways to improve healthcare, and transfer learning has shown promise in making machine learning models work better, especially when it comes to medical analysis. IoT-Enabled Transfer Learning Model (HIETLM), which is new, is proposed in this study as a way to accurately diagnose congenital heart disease. Wearable IoT devices are used by the HIETLM to continuously collect vital data from patients, such as their heart rate, blood pressure, and level of exercise. These pieces of information are sent to a central computer to be processed and analyzed. The suggested model has two main parts: a base model that was learned on a big external dataset, and a transfer learning part that was fine-tuned using data from a single patient. The base model has already been trained on a big set of general health data to learn general trends and traits that are useful for diagnosing CHD. This training helps the model understand complicated patterns and connections that are hard to learn from a small collection. The transfer learning part then tweaks the base model using the patient-specific data to make it fit the unique traits of each patient. To test how well the HIETLM worked, we did tests with a group of patients who were thought to have congenital heart disease. From the data, we can see that the HIETLM is more accurate, sensitive, and detailed than other models. The suggested model could help make CHD diagnoses more accurate and faster, which would improve patient results and lower healthcare costs.