¹D4L data4life gGmbH, Potsdam, Germany
²Hasso-Plattner-Institut für Digital Engineering gGmbH, Potsdam, Germany

Text

Introduction: Wearable and sensor data hold a promise to enhance real-time monitoring and patient engagement. The widespread use of wearable data in clinical settings is however still limited due to challenges with data interoperability, data quality, scalability, and privacy.

To address these challenges, we present the SensorHub project in which we built the D4L Collect platform that aims to facilitate the integration, harmonization, and analysis of wearable sensor data. The project is a collaboration between Data4Life and the Hasso Plattner Institute.

State of the art: The value of sensor and wearable data especially in combination with patient-reported outcomes (PROs) has been demonstrated [1]. Challenges with data privacy, security, and proprietary device software are typically addressed individually, hampering scalability. Only few frameworks have been developed to integrate data from several devices and combine it with PROs [2], [3], [4].

Concept: D4L Collect enables researchers to design, conduct, and manage remote studies. It streamlines the simultaneous collection of patient-reported and raw data from wearable and sensor devices into a harmonized dataset. Direct Bluetooth integration and encrypted data transmission ensure secure, GDPR-compliant acquisition of sensor data.

Implementation: The platform consists of two parts: Firstly, a web application with tools for project oversight by study personnel including a no-code configuration wizard for study setup, control over enrollment and consent management, and real-time adherence summaries. It offers configuration options for collection devices, and facilitates the creation of digital questionnaires (ePROs) with customizable scheduling. Data can be exported in interoperable formats for analysis.

Secondly, a mobile application serves as a single interface for study participants. It streamlines the integration of devices, via Bluetooth Low Energy (BLE), and facilitates the simultaneous collection of raw data. Participants receive targeted interactions through push notifications and data reports, and can complete ePROs and standardized assessments outside clinical settings. The app provides an in-app task manager to guide participants through assigned activities, and monitors and resolves connectivity issues with utilized devices.

Participation is possible without providing personally identifiable information (PII). Random identifiers are used for data association. Mapping between application- and clinical data is handled externally by the study team. Collected data is stored on the mobile phone and with internet connection uploaded to on-premise servers. The data is encrypted both at rest and in transit. Authorized staff can interact with and download the data through the web application, with all data access logged for comprehensive audit trails.

The platform is currently leveraged in two studies, one in stroke rehabilitation [5] and one in obsessive-compulsive disorder research. Through smartwatches, CGM sensors, and inertial measurement units participants can collect over 20 endpoints including heart rate, skin temperature, and human kinematics.

Lessons learned: The relatively smooth ethics approval process for the two studies mentioned highlights the benefits of hosting the solution and storing data on-premise at the healthcare provider where patients are treated.

For remote studies, features to detect issues (e.g., real-time adherence summaries) and support participants (e.g., in-app support for connectivity issues) are crucial for successful execution.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.

Literatur

[1] Jafleh EA, Alnaqbi FA, Almaeeni HA, Faqeeh S, Alzaabi MA, Al Zaman K. The Role of Wearable Devices in Chronic Disease Monitoring and Patient Care: A Comprehensive Review. Cureus. 2024 Sep 8;16(9):e68921. DOI: 10.7759/cureus.68921.
[2] Köhler C, Bartschke A, Fürstenau D, Schaaf T, Salgado-Baez E. The Value of Smartwatches in the Health Care Sector for Monitoring, Nudging, and Predicting: Viewpoint on 25 Years of Research. J Med Internet Res. 2024;26:e58936. DOI: 10.2196/58936.
[3] Kheirkhahan M, Nair S, Davoudi A, Rashidi P, Wanigatunga AA, Corbett DB, Mendoza T, Manini TM, Ranka S. A smartwatch-based framework for real-time and online assessment and mobility monitoring. Journal of biomedical informatics. 2019 Jan 1;89:29-40.  DOI: 10.1016/j.jbi.2018.11.003
[4] Chromik J, Kirsten K,  Herdick A, Kappattanavr A, Arnrich B. SensorHub: Multimodal Sensing in Real-Life Enables Home-Based Studies. Sensors. 2022;22:408. DOI: 10.3390/s22010408
[5] Albert J, Zhou L, Kirsten K, Kaynak N, Rackoll T, Walz T, Weese D, Kos R, Nave AH, Arnrich B. Using Wearable Sensors in Stroke Rehabilitation. In: Sensor-Based Activity Recognition and Artificial Intelligence. Proceedings 9th International Workshop iWOAR 2024; 2024 Sep 26-27; Potsdam, Germany. Cham: Springer Nature Switzerland; 2024. p. 277-282. DOI: 10.1007/978-3-031-80856-2_19