Within this framework, the goal of the motor system is to optimize some statistic of the movement such as minimizing the endpoint variance. An optimal movement is one that selleck inhibitor minimizes the deleterious effects of noise while subject to boundary constraints such
as reaching a target (on average) in a specified time. This optimization was able to predict movement trajectories for both the eye and arm (Harris and Wolpert, 1998 and Haruno and Wolpert, 2005). The benefit of this model is that the cost, i.e., accuracy, is a natural variable the sensorimotor system should care about. The cost is easy to measure because it is just how far away the hand or eye ends up from a target. The model can deal with redundancy because the noise is at the muscular level, so effects on task performance take into account the kinematics of the body
(Haruno and Wolpert, 2005). Finally, any task can be placed within the framework of optimizing the statistics of movement. For example the optimal tennis serve can be specified as the movement that has the highest probability of winning the point or has the highest speed at a particular average location with a variance mTOR cancer that means it has a 90% chance of being within the service area. However, the solutions obtained for this model were feedforward, and the incorporation of feedback required a major extension to the model. OFC was developed as a model that combined ideas on optimization with feedback control tuned to task demands (Todorov and Jordan, 2002). OFC finds the best possible feedback control law for a given task that minimizes a mixed cost function with components that specify both accuracy and energetic
costs. Subject to the dynamics of the task and the noise in the sensory and motor system, OFC finds a particular feedback control law, in other words, how particular feedback gains change throughout the movement, such that the minimal expected cost is achieved. In contrast to inverse models that map desired state and current state into a motor command, OFC does not need to specify Histone demethylase a desired state at each point in time. Instead, given a cost function that specifies a penalty on, for example, the state at some fixed time and the integrated effort, it uses the current state as an input to generate the motor command. Therefore, an important feature of these feedback control laws is that they will only correct for deviations that are task relevant and will allow variation in task-irrelevant deviations—the so-called minimum intervention principles. This matches studies that show that feedback does not always act to return the system back to the unperturbed trajectory but often acts in a manner to reduce the effect of the disturbance on the achievement of the task goal (Kurtzer et al., 2009). OFC is important as a framework because it combines trajectory generation, noise, and motor cost within a single framework and provides a clear comparison for the results of experimental work.