A clustering-based method for estimating pennation angle from B-mode ultrasound images

Abstract B-mode ultrasound (US) is often used to noninvasively measure skeletal muscle architecture, which contains human intent information. Extracted features from images can help improve closed-loop human–robotic interaction control when using rehabilitation/assistive devices. The traditional manual approach inferring the structural US laborious, time-consuming, and subjective among different investigators. This paper proposes a clustering-based detection method that mimic well-trained expert in identifying fascicle aponeurosis and, therefore, compute pennation angle. architecture assumes fibers have tubular characteristics. It robust for low-frequency image streams. We compared proposed algorithm two mature benchmark techniques: UltraTrack ImageJ. performance of showed higher accuracy our dataset (frame frequency 20 Hz), is, similar expert. shows promising potential automatic orientation facilitate implementations biomechanics modeling, rehabilitation robot design, neuromuscular disease diagnosis with data stream.