Predictive analytics

The main goal is to create a predictive algorithm that can predict the subcutaneous glucose concentration for individuals with type 1 diabetes. Data for the algorithm is collected from various sources, including mobile devices such as wrist monitors, smartphones, and continuous glucose monitoring devices. The data encompasses a range of variables, such as subcutaneous glucose concentration, dietary information, insulin doses, and vital signs. The collected data is transmitted via Bluetooth to a mobile app, which serves as a gateway for data collection. This data is then stored in private cloud databases. The core of the service involves developing a machine learning algorithm that can analyze the collected data to predict glucose concentration values for the next 15 minutes, half an hour, and one hour. Once the algorithm makes these predictions, it triggers alerts to notify the user about the projected glucose concentrations.