Digital Measures
Catalogue
All measures start with the reliable collection of sensor data and the delivery of those data into our infrastructure.
This sensor data includes movement, acceleration, light, temperature, smartphone interactions and many more. There are many approaches available to process these signals.
For example, when moving, the output of an accelerometer can be combined to understand the energy or intensity associated with the movement. When still, information about posture is also available due to the constant reference of earth’s gravity.
Defining Measures
The sensor-level (raw) data output of the GENEActiv allows these event measures to be defined and calculated offline, after data collection. While the ActivInsights Band uses on-board algorithms to detect and characterise sensor data in real-time, transmitting wirelessly to a secure global data infrastructure.
The event measures from either of our wearables and other connected devices can then be built up into daily summary measures and longer-term measures. This hierarchy of measures forms the basis of the ActivInsights Digital Measures Catalogue with the measurements from remotely deployed, digital sensors as the foundation.
Each measure, within every layer of the hierarchy is independently verified and validated in our data infrastructure and referenced to peer-reviewed publications.
We recommend digital health measures according to your research objectives in different therapeutic areas.
Contact us to find out more about the 150+ health and clinical measures that we regularly use from our Digital Measures Catalogue, including biomarker development and exploratory endpoints.
Example Measures Include:
Example Measures Include:
Analysing Signals
Analysis of signals in the frequency domain provides information about repeated movements. This could include scratching, tremor and environmental interactions. Repetitive or pre-defined movement patterns can also be detected.
In the first stage of analysis, periods of time where the signal is consistent (homogenous) are defined as an ‘event’, typically measured in seconds or minutes. The properties of the signal are used to characterise and then classify the event, for example, a 90 second walking bout at an average cadence of 74 steps per minute with a total of 111 steps.
