A Pilot Study on Quantitative Muscle Strength Assessment Using EMG and Motion Data

Assessing muscle strength is an important aspect of evaluating motor function. While clinical scales are commonly used due to their simplicity and practicality, they often rely heavily on subjective assessments by clinicians and lack quantitative precision. This paper introduces a method for quantitative muscle strength assessment using a wearable electromyography (EMG) sensor and inertial measurement units (IMUs). Following the linguistic guidelines of the Lovett Muscle Strength Scale, we developed an algorithm based on fuzzy logic for muscle strength classification. A pilot study involving fifteen patients with spinal cord injurie (SCI) demonstrated that the muscle strength grades obtained through our proposed method exhibit a high correlation with assessments made by clinicians (r = 0.966∼0.977), with root mean square errors ranging from 0.27 to 0.36. This algorithm is appropriate for clinical application as it is not machine learning based method, thereby eliminating the need for extensive training on vast sets of clinical data.