Wearable Sensors: A Valid Tool for Quantifying Cervical Spondylotic Myelopathy

Introduction: Though nearly ubiquitous in testing for cervical spondylotic myelopathy (CSM), the value of the conventional Romberg test is constrained by its binary nature. Mild cases of myelopathy are often managed nonoperatively. However, differentiating which patients have mild cases from those who require surgery more urgently is challenging without objective metrics. The recent advance of performing a Romberg test on a force plate enables a more granular measure of imbalance in patients with CSM. Nonetheless, the use of force plates limits the amount of patient data that can be collected and the setting in which they can be collected. The advent of wearable sensors offers the opportunity to measure imbalance in patients when they are away from the clinical setting. However, this has not yet been validated against another instrument such as the quantitative Romberg test.

Methods: 45 subjects with CSM scheduled for surgery underwent Romberg testing using force plates and wearable sensors (placed at the C7 level). Data on force plate displacement (measured in mm of displacement) was compared to motion data from the wearable sensors (measured in degrees of angular displacement).

Results: There was a strong correlation between data output from the force plate (the “quantitative Romberg” test) and the data output from the wearable sensors—suggesting the feasibility of using sensors to quantify CSM severity. When eyes are closed preoperatively for the Romberg test, there was a strong correlation between the wearable sensors data and the quantitative Romberg data from the force plate when measuring maximum lateral motion (r=0.76), and when measuring total lateral center of pressure (COP) motion (r=0.68). Similarly, there was a moderate correlation between the sensors and the force plate when performing the eyes closed Romberg test and measuring total COP motion (r=0.58).

Conclusion : Wearable sensors present a growing subset of remote digital health technology to gather biomechanical gait and stance data. These data can elucidate the disease course and manifestations of conditions like CSM and may drive diagnostic and therapeutic decisions in the future.