Little is known about the specific role of somatosensors in multijoint posture control. While heteronymous feedback loops are known to exist, it is unclear what their specific role and importance is in balance control. Vestibular and visual information is also known to interact with somatosensory feedback for posture control, but the specific interaction is unknown. Theoretical models can help us study sensorimotor integration. Models have demonstrated that body dynamics constrain CNS control strategies, and that spatial orientation perception can be modelled by an internal representation of the body used to form an optimal estimate of body orientation. We hypothesize that homonymous feedback of somatosensory information is insufficient to properly maintain balance, and that somatosensory information is integrated with vestibular and visual information in an optimal estimate of body state, similar to that shown in perception studies. We propose to use computational models of body, muscle, and sensory dynamics, based on experimental data, to test these hypotheses. The long-term goal of this study is to gain an increased understanding of the integration of visual, vestibular, and somatosensory information in human posture control. Such understanding is necessary for the development of intervention and rehabilitation measures for persons suffering from sensory dysfunction.
The specific aims of this study are: 1. To develop a model of body somatosensors to augment existing models of body biomechanics. To supplement current work on vestibular models to provide a full-body model of human posture control. Parameters of this model are to be derived from existing experimental data. 2. To utilize the model to determine whether homonymous feedback of somatosensory information is sufficient for proper posture control. To perform computer simulations in which motor control is constrained to use only homonymous feedback, and to assess the stability of the that system. The tests are to be compared with multiple trials from concurrently funded experiments. 3. To utilize the model to determine whether somatosensory and vestibular information is integrated in an optimal estimation of the body state used for posture control. To predict the covariance of ankle-hip movement for upright standing, under sensory conflict conditions. The model is to be compared directly with experimental covariances for conflicting visual and somatosensory information, as measured using a standard clinical balance test. The project described here takes advantage of experimental results and model development funded by a newly established NIDCD/NASA Center for Vestibular Research. The project emphasizes collaboration with other investigators to provide a computational modelling complement to experimental studies.
