Project GRIP
We use machine learning to translate motion sensor data into activity levels. By combining the sensor and algorithm, we analyze the activity patterns of a large group of individuals with chronic pain. Additionally, we examine variations within this group, such as differences between individuals with severe or mild pain and comparisons with those without chronic pain.
Objectives
- Processing IMU Data: Develop an algorithm to process raw sensor data into activity levels.
- Transforming Activity Levels: Create algorithms to translate activity levels into meaningful characteristics, such as minutes of inactivity.
- Analytics: Evaluate the clinimetric properties of the activity characteristics.
- Patterns: Explore associations between pain, activity levels, and other characteristics.
Timeline
| Milestone | Expected Completion |
|---|---|
| Data processing | Q1–Q3 2024 |
| Data transformation | Q3–Q4 2024 |
| Analytics | Q1 2025 |
| Publications | 2025–2027 |
Team Members
- Annet Doomen - Project Manager
- Richard Felius - Lead Developer
Current Status
The machine learning algorithms have been tested and validated. An algorithm has been developed to process the results into clinically relevant insights, such as sedentary time. The findings will be published between 2025 and 2026.
Resources
Available after publication.
Contact
For more information, please contact Richard Felius at Richard.felius@hu.nl.