Spectral Analysis Based Limb Movement Reconstruction Using Inertial Sensing Data

This brief paper illustrates the analysis of inertial sensing signals including linear acceleration and angular velocity for the reconstruction of movement trajectory of upper limbs, in order to assess the performance of stroke rehabilitation and build up an objective quantitative evaluation system. It is well noted that while using inertial sensing data, the motion-track reconstruction applies the Kalman filtering to process integrated time waveform [1–3]. This study consists of two parts. First, the method of track reconstruction of limb motion is built up through simulation. We present the new method to obtain angular displacement using spectral analysis by extracting the amplitude and phase of significant frequency components that correspond to limb movement. Then, the upper- and fore-arm positions corresponding to global coordinate system can be computed through the quaternion vector. Secondly, the proposed technique is verified using five designated upper-limb movements including upper limb rotating (1) upwards 90...