A framework for harmonizing sensor data to support embedded health assessment.

Area of Science:

• Digital Health
• Biomedical Engineering
• Health Informatics

Background:

• In-home and mobile sensing are crucial for future healthcare, aiding in early illness detection and chronic condition management.
• Current research collaboration and data sharing are hindered by diverse sensors and data collection methods.
• Lack of standardized measures across different sensor types limits large-scale studies and dataset reuse.

Purpose of the Study:

• To propose a framework for harmonizing heterogeneous sensor data.
• To enable consistent data analysis across multiple research sites and sensor types.
• To facilitate large-scale health studies and the reuse of existing datasets.

Main Methods:

• Development of a framework utilizing an intermediate layer, the Conceptual Sensor.
• Mapping of physical measures from heterogeneous sensors to a standardized clinical space.
• Demonstration with examples for sleep quality and ambulatory physical function.

Main Results:

• The proposed framework enables harmonization of data from disparate sensors.
• The Conceptual Sensor acts as a bridge between physical sensor measures and clinical interpretation.
• Successful application demonstrated for sleep quality and ambulatory physical function measures.

Conclusions:

• A novel framework for harmonizing heterogeneous sensor data is presented.
• The Conceptual Sensor approach facilitates cross-site research and data integration.
• This work supports the advancement of large-scale digital health studies and data-driven healthcare.