Ear-worn IMUs can extract a broad range of sit-to-stand and sit-to-walk outcome measures, with temporal outcomes demonstrating the closest agreement with optical motion capture, while kinematic, power, and fluency measures showed varying degrees of systematic and proportional disagreement.
Key Findings
Results
Temporal outcomes showed the closest agreement between ear-worn IMU and optical motion capture systems for both sit-to-stand and sit-to-walk movements.
Mean overall movement duration differed by 0.09 (SD 0.12) seconds for sit-to-stand and 0.06 (SD 0.17) seconds for sit-to-walk.
95% limits of agreement were -0.14 to 0.32 seconds for sit-to-stand and -0.28 to 0.39 seconds for sit-to-walk.
Agreement was evaluated using linear mixed-effects analyses accounting for repeated trials within participants.
Outcome metrics included overall and subphase durations, acceleration, velocity, power, movement smoothness, and hesitation.
Results
Peak power showed strong linear relationships between ear-worn IMU and optical reference but was not interchangeable due to substantial systematic bias.
Mixed-effects regression slopes were 0.70 for sit-to-stand and 0.68 for sit-to-walk, indicating strong linear relationships.
Systematic biases of 172 (95% CI 153-190) W for sit-to-stand and 191 (95% CI 168-215) W for sit-to-walk were observed.
The magnitude of these biases indicated that ear-derived and seventh cervical vertebra-derived power were not interchangeable.
Results
Peak upward acceleration showed systematic underestimation by the ear-worn IMU compared to the optical reference for both movements.
Systematic bias for peak upward acceleration was 1.05 (95% CI 0.93-1.18) ms-2 for sit-to-stand.
Systematic bias for peak upward acceleration was 1.30 (95% CI 1.14-1.45) ms-2 for sit-to-walk.
These biases indicate that ear-derived acceleration measurements were consistently lower than those from the seventh cervical vertebra optical marker.
Results
Hesitation demonstrated comparatively small absolute error, while smoothness and other fluency outcomes showed greater disagreement between measurement systems.
Hesitation root mean squared error was 0.04 for sit-to-stand and 0.03 for sit-to-walk.
Smoothness and other fluency outcomes showed greater disagreement between the ear-worn IMU and optical reference.
Agreement was evaluated using systematic and proportional bias, limits of agreement, root mean squared error, and smallest detectable difference.
Methods
The study enrolled 62 participants including both healthy individuals and those with known movement conditions, using a bespoke 3D-printed headset with bilateral ear-placed 6-axis IMUs.
46 participants did not have a known underlying movement condition (18 male, 28 female; median age 27, IQR 15 years).
16 participants had a known underlying movement condition (7 male, 9 female; median age 62, IQR 23 years).
Each participant performed 3 sit-to-stand movements and 3 sit-to-walk movements.
A reflective marker was placed on each participant's seventh cervical vertebra to provide the optical motion capture reference.
The headset contained 6-axis IMUs placed beside the left and right ears.
Conclusions
The study established feasibility of using ear-worn IMUs for comprehensive functional movement assessment, supporting subsequent validation in clinically relevant populations.
Ear-worn IMUs extracted a broad range of sit-to-stand and sit-to-walk outcome measures.
The findings provide a basis for subsequent validation of the most robust outcomes in clinically relevant populations.
The approach is noted as relevant to monitoring older adults and individuals with movement disorders with minimal additional burden, as IMUs are increasingly incorporated into hearing aids and headphones.
What This Means
This research suggests that sensors worn near the ears — similar to those already found in hearing aids and wireless earbuds — can measure how people rise from a chair and begin walking. The study tested a specially designed headset containing motion sensors on 62 people, including both healthy younger adults and older adults with movement conditions, comparing the headset's measurements against a gold-standard optical motion capture system. The results showed that timing measurements (how long each phase of movement takes) were the most accurate, differing from the reference system by only about a tenth of a second on average, while measurements of power and acceleration were consistently lower from the ear sensors compared to the neck-based reference, meaning the two systems cannot simply be swapped for one another without adjustment.
This research suggests that the location of the sensor (the ear versus the base of the neck) matters for some types of measurements but not others. For capturing the smoothness and fluency of movement, the ear-worn sensors were less reliable, but for detecting hesitation — brief pauses before movement — they performed well. The practical implication is that ear-worn devices could one day allow clinicians or researchers to monitor functional movement in everyday life without requiring patients to visit a specialist laboratory or wear additional equipment beyond what they might already use.
This research suggests a path toward using consumer technology like hearing aids or headphones as passive health monitoring tools for older adults or people with movement disorders, though further validation in these specific clinical populations is needed before the findings can be applied in real-world care settings. The study identifies which types of measurements are most trustworthy from ear-worn sensors, providing a practical roadmap for future research.
Check Your Own Numbers
Upload your bloodwork. We'll cross-reference your results against this study and 4,700 others.
Fawden T, Edwards A, Roberts I, Dunne H, Stone T, Bance M. (2026). Ear-Worn Inertial Measurement of Sit-to-Stand and Sit-to-Walk Movements: Algorithm Development and Validation.. Journal of medical Internet research. https://doi.org/10.2196/100936