Abstract: Quantitative evaluation on the level of hemiparesis for post-stroke patients is still challenging. In this study, we proposed an innovative method for investigating in vivo muscle behavior in hemiparetic gait. A wearable data collection platform was developed for the simultaneous acquisition of ultrasonography (US), electromyography (EMG) and joint angle during gait cycles. Using this platform, we studied twelve hemiparetic patients during walking on a treadmill and analyzed the corresponding characteristics of their medial gastrocnemius (GM) muscle. Several interesting outputs are achieved such as the observation that dynamic ranges of the both the pennation angle (PA) of GM and EMG signal from same side were more diffused, for the affected side than for the unaffected side. The results suggested that the proposed multimodal signal acquisition and wearable platform is capable of providing comprehensive in vivo information regarding dynamics of ankle and medial gastrocnemius. The proposed platform could be a promising tool for the assessment of hemiparesis patients.