There are two use cases for Everion in research, beyond simply using Everion as a reference device in a study.
- validate Everion in specific contexts, to explore the performance of the Everion in specific contexts against a reference measurement
- Collect data. You can either draw on our own data collection ("test bench") to see underlying performance in specific contexts, or you can collect your own data set. Here, we provide tools, a recommendation and checklist for how to do this efficiently.
- Process automatically. If you collect your own data, we provide an automated stand-alone software to convert Everion signals (e.g., heart rate) and reference signals (e.g., ECG) to be ready for analysis. This includes the ability to interactively annotate data. If you use other reference devices than ours, we can build parsers to process them too.
- Analyse interactively. We provide you with a tool to interactively explore and chart performance of Everion signals against reference. To the extent that we can share data, we are able to adapt our algorithms to better capture your specific contexts
- Develop algo. We provide a Matlab environment for you to create your own function that can draw on all Everion sensor* and signal data for calculation, the results of which can be explored interactively.
- Deploy algo on Everion. We offer a service to compile your Matlab code to deploy it on the Everion natively. Deployment is subject to constraints, such that code needs to be compilable, be sufficiently compact to fit onto the Everion and stay within certain computational limits.
*sensor data are what is generated by the sensors on the Everion, e.g.,
- accelerometer data at 50Hz in three directions
- LED data for three wavelengths at 51.2Hz
- 3 temperatures (skin, board, other) at 1 Hz
- 1 barometer reading at 1 Hz
- 1 Electrodermal activity reading at 1-10 Hz
signal data, by contrast, are the outputs of the Everion, including Heart rate, respiration rate, blood oxygen saturation, etc.