Towards an extensible ontology for streaming sensor data for clinical trials

ABSTRACTThe use of wearable sensors for clinical trials can lead to better data collection and a better patient experience during trials, and can further allow more patients to participate in trials by allowing more remote monitoring and fewer site visits. However, extracting maximum value from the data collected via streaming sensors presents some specific technical challenges, including processing the data in real time, and storing the sensor data in a representation that facilitates the use of biomarker algorithms that can be used and reused with different similar sensors, at different scales, and across different clinical trials. Here we present our initial work on SORBET, a Sensor Ontology for Reusable Biometric Expressions and Transformations. Our design strategy is presented, along with the initial design and examples. While this ontology has been created for the Medidata Sensor Cloud product, it is our hope that others working in this space will join us in extending and hardening this ontology, as we expand it to incorporate more sensors and more needs for clinical trials research.