Poster Y10 - Surface Acoustic Wave (SAW) Force Myography Sensor for Muscle Force Output Identification

Introduction: Monitoring limb movements and muscle activity is crucial in rehabilitation, sports science, and wellness, as understanding biomechanics is key for assessing physical performance and customizing training or therapy. Advancements in wearable technology have led to innovative methods for monitoring muscle activity, notably through surface electromyography (sEMG) and force myography (FMG). While sEMG measures electrical signals from muscle contractions and faces challenges with noise interference, FMG offers a non-invasive alternative by tracking mechanical deformation in muscles for force estimation. FMG's benefits include cost-effectiveness and a superior signal-to-noise ratio, making it a promising option for accurate muscle activity assessment. Surface Acoustic Wave (SAW) technology, known for its application in strain sensing and passive, wireless operation, has seen recent advancements, making it increasingly suitable for wearable technologies. Its design flexibility allows for highly sensitive and thermally stable strain sensing. However, despite its promising potential for muscle force monitoring, the use of SAW sensors in this context remains unexplored. This study introduces a novel application of SAW sensor technology within the category of FMG for accurately detecting muscle contractions and relaxations, particularly during upper arm exercises like bicep curls.

Materials and

Methods: The fabrication of the SAW sensors involved several steps, starting with 128° YX-cut LiNbO₃ wafers, which were cut, cleaned, and then bonded to copper bars. Interdigitated transducers (IDTs) were created using chromium (Cr) and gold (Au) through a lift-off process involving multiple photoresist layers. Detailed patterning was achieved with a JEOL e-beam writer (JEOL JBX-6300FS), followed by the development and deposition processes to form the sensor's active elements. The completed SAW sensor was then integrated into an adjustable armband, connecting to SMA connectors for data transmission to the network analyzer (E5061B-005, Keysight Technologies) (Fig. 1). The approach to measuring muscle activity in the biceps brachii during preacher curl exercises involved implementing two distinct testing protocols: static and dynamic. For static tests, the sensor captured data on muscle contractions at varying weights (0 lb-20 lb), with phase shifts analyzed to assess strain, complemented by strain gauge (Tokyo Measuring Instruments Lab, model CFLA-3-350) and electromyography (EMG) sensor (MyoWare Muscle Sensor AT-04-001) comparisons for validation. Dynamic testing involved repeated curl motions with different weights (0 lb-20 lb), collecting data simultaneously from the SAW sensor, strain gauge, and EMG sensor to evaluate muscle activity in real-time. Data processing included bandpass filtering of EMG signals and normalization of SAW sensor phase shifts against established baselines. Pearson's correlation analysis was conducted across the datasets to examine the consistency and accuracy of muscle activity measurements.

Results, Conclusions, and Discussions: The SAW sensor was engineered to achieve its resonant frequency at 915 MHz when the magnitude of the S_11 parameter is at its lowest, achieving this in practice. It exhibited a Strain Coefficient of Delay (SCD) of −0.418 ppm/με. In static tests, the sensor's response to varying loads mirrored that of a traditional strain gauge (Fig. 2). A phase shift of -0.65 × 10^(-3) rad/λ was observed under a load increase from 0 to 5 lbs, corresponding to a microstrain shift of −314 με, indicating significant muscle expansion. However, for a load increase from 10 to 20 lbs, the phase shift reduced to -0.2 × 10^(-3) rad/λ with a microstrain change of −119με. For the dynamic test, the SAW sensor's outputs showed a strong positive correlation with the strain gauge (r=0.790, p< 0.001) and a moderate inverse relationship with the EMG gauge outputs (r=−0.550, p< 0.001). Similarly, the strain gauge and EMG outputs had a moderate negative correlation (r=−0.569, p< 0.001), underscoring the consistent inverse trend between muscle strain and electrical activity measurements. This research introduced the novel use of SAW sensor technology for FMG to monitor muscle contractions and relaxations. The setup offered quantitative insights into muscle force output under various static and dynamic conditions. The findings suggest that SAW sensors may offer a promising alternative for capturing muscle dynamics, potentially outperforming traditional methods such as strain gauges and EMG, indicating their potential to enhance muscle activity detection.

Acknowledgements (Optional):