eHealth for Personalized Monitoring and Collective Analysis.
People with mild conditions, more serious chronic ones, or in rehabilitation, are being remotely monitored and advised in their common, everyday settings by medical professionals. People themselves monitor also their quality of life, keeping track of adverse events like migraines or athletic injuries and the effect of those on their lifestyle. Additionally, professionals in para-medical domains (wellness, fitness, nutrition) monitor and coach their clientele.
The aforementioned situations have in common the (self) monitoring of people for long periods of time in their everyday life, for collecting and analysing real-world data (RWD) in order to coach them. At the cornerstone of this coaching service are three fundamental ML services: (i) The inference about a patient given some of their data and a predictive model. (ii) The phenotyping, i.e. the assignment of a patient to a most descriptive of a set of generative models given some of their data. Finally, (iii) the synthesis of fully anonymous data given a set of generative models and a model transition probability matrix, trained from the behaviour of the patient.
Use Case Objectives
Use Case Leader
Innovation Sprint (iSPRINT) is a high-tech SME involved in a variety of eHealth services. The company applies Signal Processing and Machine Learning/ Artificial Intelligence techniques on signals collected from IoT devices to better understand Physical Data and link them to clinical outcomes and drug effectiveness. These algorithms are then integrated into systems encompassing both software and hardware, building systems that observe and understand. iSPRINT is adapting algorithms and models from its Healthentia eClinical platform and is involved in all aspects of the eHealth use case: the definition of its different scenarios (inference, phenotyping and data synthesis), the experimentation leading to the necessary FaaS implementations and the evaluation.