Berlin 2015 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

BP: Fachverband Biologische Physik

BP 37: Modelling of non-linear dynamics in biological movement (focus session)

BP 37.1: Invited Talk

Wednesday, March 18, 2015, 15:00–15:30, EW 202

The cost of moving optimally — •Dinant Kistemaker — VU, Amsterdam, The Netherlands — UWO, London, Canada.

The field of Human Motor Control is concerned with how the brain deals with the very many kinematic and mechanical degrees of freedom (DoF) to control posture and movement. In this field, mathematical models of the musculoskeletal system are indispensable as they provide answers to questions that are inaccessible by experimental studies alone. In a recent set of studies, predictions using a detailed model of the arm together with behavioural data were used to investigate if the DoF are exploited by the brain to minimize costs at three distinct levels: the motor system's input (e.g. control effort), the motor system's mechanical output (e.g. energy) and kinematics (e.g. jerk). Subjects performed goal-directed arm movements while holding on to a robotic manipulandum in combination with visual perturbations of seen hand position. The force fields created by the robot and visual perturbations were specially designed to be able to independently change the costs at the three levels. Direct Collocation was used to translate the ODE's of the model into nonlinear constraints and were solved together with task and boundary constraints using SNOPT, while minimizing several costs at the three levels. It was found that the behavioural data was inconsistent with the notion that the brain minimizes energy expenditure. Furthermore, it was found that in selecting a kinematic path, the brain does not take into account costs that relate to the input level or the dynamic level. Movement patterns observed experimentally were only consistent with a cost function based solely on kinematic costs.