A postural control model incorporating multisensory inputs for maintaining a musculoskeletal model in a stance posture. (17th January 2017)

• Record Type: Journal Article
• Authors: Jiang, Ping; Chiba, Ryosuke; Takakusaki, Kaoru; Ota, Jun
• Is Part Of: Advanced robotics. Volume 31:Number 1/2(2017)
• Page Start: 55
• Page End: 67
• Publication Date: 2017-01-17
• Subjects: Musculoskeletal simulation -- stance postural control -- muscle tonus -- forward dynamics simulation -- sensory integration; Robotics -- Periodicals; Robotics -- Japan -- Periodicals; Robotics; Japan; Periodicals; 629.89205
• DOI: 10.1080/01691864.2016.1266095
• Languages: English
• ISSNs: 0169-1864

Abstract

Graphical Abstract: Abstract : Maintenance of the upright stance is one of the basic requirements in human daily life. Stance postural control is achieved through muscle coordination based on the integration of multisensory inputs such as visual, vestibular, and proprioceptive somatosensory inputs. In this study, our aim was to develop a stance postural control model including a neural controller with feed-forward control and sensory feedback control based on visual, vestibular, and proprioceptive somatosensory feedback inputs. The feed-forward control consists of conventional feed-forward control, which sends constant inputs (necessary to maintain the posture) to the muscles, as well as stiffness regulation which controls the body stiffness (overall joint stiffness) based on sensory input conditions. In contrast to previous studies, a musculoskeletal model with 70 muscles rather than an inverted pendulum was selected to represent the human body. A bipedal stance under the influence of a maximal 120-ms neurological time delay was achieved by muscle control rather than joint torque control. Furthermore, we considered four sensory input conditions: normal, vision exclusion, vestibular loss, and vision exclusion together with vestibular loss. In each condition, variables of the neural controller were searched to keep the musculoskeletal model standing during a five-second forward dynamics simulation by the optimization method. As a result, we found that when all three types (visual, vestibular, and proprioceptive) of the sensory inputs are available, low muscle stiffness is sufficient to maintain the balance of a musculoskeletal model. When one (visual or vestibular) or several (both visual and vestibular) sensory inputs get excluded, high stiffness may be desirable for maintenance of the balance. This simulation result indicates that our newly developed computational model resembles the physiological features reported in experimental study, namely that individuals might stiffen their body during upright standing to cope with deterioration in sensory input.