IoT and Machine Learning approach for Early Heart disease Prediction & Diagnosis

Heart disease is one of the most prominent causes of deaths globally. Every year almost 17.9 million people lose their life due to heart disease which account for 31% of total deaths worldwide. Most of the time patients know about their heart disease after reaching a severe heart condition from where total recovery is impossible. However, if the Cardiovascular state is monitored regularly, heart disease can be detected at an early stage and early detection can prevent the severity of most heart diseases. In most cases patients do not feel any kind of pain when the cardiovascular diseases grow slowly. By the time someone feels uneasiness and pain, their heart condition gets seriously bad. Moreover, it is also not feasible for everyone to check up on their heart condition periodically by visiting a heart specialist. Our proposed system will work for the early detection of heart diseases using Machine learning classifiers and IoT technologies. Our system has two subsystems. First one is our trained machine learning model which will be implemented as a WebApi. Second one is our IoT setup with heartbeat sensors. Sensors will collect data from the user's body and send those to the machine learning model. Then, the model will predict the result about the user’s heart condition and send it back to the IoT device. Model will classify the user’s heart condition either as “Normal” or “Abnormal”. Based on the result, the user should go to a cardiologist for a checkup. We have used the Heart Disease Dataset from UCI Machine Learning Repository. In addition, we trained seven machine learning algorithms after preprocessing the dataset. Further we will also build an IoT setup with sensors to communicate with the WebApi and complete our proposed system of predicting heart diseases.