Data Analytics for Health and Connected Care: Ontology, Knowledge Graph and Applications

Connected care applications are increasingly used to achieve a more continuous and pervasive healthcare follow-up of chronic diseases. Within these applications, objective insights are collected by using Artificial Intelligence (AI) models on Internet of Things (IoT) devices in patient’s homes and by using wearable devices to capture biomedical parameters. However, to enable easy re-use of AI applications trained and designed on top of sensor data, it is important to uniformly describe the collected data and how this links to the health condition of the patient. In this paper, we propose the DAHCC (Data Analytics For Health and Connected Care) ontology, dataset and Knowledge Graph (KG). The ontology allows capturing the metadata about the sensors, the different designed AI algorithms and the health insights and their correlation to the medical condition of the patients. To showcase the use of the ontology, a large dataset of 42 participants performing daily life activities in a smart home was collected and annotated with the DAHCC ontology into a KG. Three applications using this KG are provided as inspiration on how other connected care applications can utilize DAHCC. The ontology, KG and the applications are made publicly available at https://dahcc.idlab.ugent.be . DAHCC’s goal is to integrate care systems such that their outcomes can be visualised, interpreted and acted upon without increasing the burden of healthcare professionals who rely on such systems.