JMIR Publications

ABSTRACT

Background:

As the phenomenon of population aging continues, increased pressure is placed on healthcare services to treat widespread movement disorders which disproportionately affect older adults. Sit-to-stand and sit-to-walk movements are common, universally understood daily activities which provide deep insight into an individual’s health status. Developing comprehensive analysis metrics for sit-to-stand and sit-to-walk movements based on ear-worn inertial measurement units, which are found in widely used devices such as hearing aids and headphones, unlocks a prime opportunity to improve the monitoring of elderly individuals with movement disorders with minimal burden.

Objective:

This paper aims to introduce comprehensive, clinically useful analysis metrics for sit-to-stand and sit-to-walk movements based on data from ear-worn inertial measurement units, and to evaluate them against a gold standard optical motion capture system.

Methods:

Data were collected from 61 participants, who were organised into a typically functioning group (16M/34F, median age 27 (IQR 15)) and a group comprising those suffering from an underlying movement condition (7M/9F, median age 62 (IQR 23)). The participants wore a bespoke 3D-printed headset which contained 6-axis inertial measurement units (IMUs) placed beside their left and right ears. A reflective marker was placed on each participant’s 7th cervical vertebra (C7) to provide comparison with a gold standard optical motion capture system. Each participant performed 3 sit-to-stand movements and 3 sit-to-walk movements. After calculating the correlation between the IMUs and C7 marker data, the total movement duration, duration of each subphase, peak acceleration and velocity, power, movement smoothness and hesitation were analysed offline using the IMU and C7 marker data.

Results:

Very strong correlation scores (r≥0.99) were found between the ear-worn IMUs and C7 marker for forwards and upwards acceleration during sit-to-stand and sit-to-walk. The mean overall movement durations measured using the ear-worn IMUs were within 0.10 seconds of the optical system, while very strong linear relationships were observed between the power calculated by the ear-worn IMUs and the optical system for sit-to-stand (r=0.97) and sit-to-walk (r=0.94). Strong correlations were found between the IMU and optical systems when measuring hesitation in sit-to-stand (r=0.80) and sit-to-walk (r=0.88), though were comparatively poor for movement smoothness (r=0.47 for sit-to-stand and r=0.54 for sit-to-walk respectively).

Conclusions:

The proposed sit-to-stand and sit-to-walk movement metrics, including movement duration, power and peak acceleration and velocity, can be accurately analysed using ear-worn IMUs. These metrics provide a basis for to comprehensively analyse these movements using devices which are already integrated into users’ daily routines.