ヒト軟組織による情報処理
Information processing via human soft tissue
ヒトの軟組織の非線形・粘弾性を物理リザバー計算に利用し、手首関節角度と超音波画像から筋変形を状態として非線形力学系を模倣できることを示した。
著者: Yo Kobayashi
分類: q-bio.NC, cs.AI, cs.HC, cs.LG, cs.RO, cs.SY, eess.SY
原文アブストラクト
This study demonstrates that the soft biological tissues of humans can be used as a type of soft body in physical reservoir computing. Soft biological tissues possess characteristics such as stress-strain nonlinearity and viscoelasticity that satisfy the requirements for physical reservoir computing, including nonlinearity and memory. The aim of this study was to utilize the dynamics of human soft tissues as a physical reservoir for the emulation of nonlinear dynamical systems. To demonstrate this concept, joint angle data during motion in the flexion-extension direction of the wrist joint, and ultrasound images of the muscles associated with that motion, were acquired from human participants. The input to the system was the angle of the wrist joint, while the deformation field within the muscle (obtained from ultrasound images) represented the state of the reservoir. The results indicate that the dynamics of soft tissue have a positive impact on the computational task of emulating nonlinear dynamical systems. This research suggests that the soft tissue of humans can be used as a potential computational resource.
