Wearable Active Vibration Sensing for Mid-Activity Knee Health Assessment

• Main Authors: Ozmen, Goktug C., Nichols, Christopher J., Lan, Lan, Moise, Emily, Sugino, Christopher, Erturk, Alper, Inan, Omer T.
• Format: Conference Proceeding
• Language: English
• Subjects: Atmospheric measurements; Biomedical monitoring; Frequency measurement; Injuries; joint health; Particle measurements; Sensors; Vibration measurement; Vibrations; Vibrometers; Wearable computers; wearable sensing
• ISSN: 2376-8894
• DOI: 10.1109/BSN58485.2023.10331249

Non-invasive vibration measurements from the knee offer a convenient and affordable alternative to benchtop or biomechanics lab joint health monitoring systems. Recently, joint acoustic emissions (JAEs) measured from the knee were shown to be an indicator of knee health. However, the origin of JAEs is still not fully understood, which limits its acceptance and use by clinical experts. In this proof-of-concept study, rather than relying on the movements of the knee and corresponding frictional rubbing of internal surfaces to produce vibrations, we propose using an active vibration sensing approach with a known vibration source interrogating the knee. We aim to elucidate the linkage between knee vibration characteristics and structural changes in the joint following injuries. We measured tibial vibration responses of two participants using a laser vibrometer system to quantify the frequency band where the most repeatable tibial vibration measurement can be taken. Subsequently, a custom-designed wearable system measured mid-activity tibial vibration characteristics from four participants (five healthy knees and three knees with prior acute injury) during unloaded knee flexion-extensions. An active sensing knee health score was defined as the ratio of the changes in low- to high-frequency response during flexion-extension. Since changes in the boundary of tibia would alter low-frequency response more than high frequency response, we found that increased knee laxity with acute injuries resulted in an increased active sensing knee health score. Our findings demonstrate the potential of active vibration sensing as an interpretable, computationally inexpensive alternative to JAEs for wearable knee health assessment.