Smart Gait Detection System With IoT and Machine Learning

Mild stroke symptoms and transient ischemic attack (TIA) normally lasts for sometime and it doesn’t cause any major issues. Sometimes patient is not aware of the symptoms and increases chances of balance loss and fall. These patient need sudden medical help for avoiding permanent disabilities occur due to fall and stroke. The continuous monitoring of movement of the patient can play an important role in detection of different types of gait and prediction chances of fall early. The main aim of the project is to detect gait using sensors and predict chances of fall using machine learning. The movement of patient is detected using three axis accelerometer and force on foot detected by force sensor placed below the foot. The real-time position of patient is pointed on the website with support of global positioning system (GPS) module included in the system. IoT uses in the system for updating sensor data to cloud server (Thingspeak) for conducting study on the data. Machine learning algorithm is applied on the acquired sensor data for predicting the chances of fall. The result shows around 95% accuracy for super vector machine (SVM) algorithm we used for training and prediction.