Intercepting and avoiding collisions with moving objects are fundamental skills in daily life. Anticipatory behavior is required because of significant delays in transforming sensory information about target and body motion into a timed motor response. The ability to predict the kinematics and kinetics of interception or avoidance hundreds of milliseconds before the event may depend on several different sources of information and on different strategies of sensory-motor coordination. What are exactly the sources of spatio-temporal information and what are the control strategies remain controversial issues. Indeed, these topics have been the battlefield of contrasting views on how the brain interprets visual information to guide movement. Here we attempt a synthetic overview of the vast literature on interception. We discuss in detail the behavioral and neurophysiological aspects of interception of targets falling under gravity, as this topic has received special attention in recent years. We show that visual cues alone are insufficient to predict the time and place of interception or avoidance, and they need to be supplemented by prior knowledge (or internal models) about several features of the dynamic interaction with the moving object.

McIntyre. Anticipating the effects of gravity when intercepting moving objects: differentiating up and down based on nonvisual cues. J Neurophysiol 94: [4471][4472][4473][4474][4475][4476][4477][4478][4479][4480] 2005. First published August 24, 2005; doi:10.1152/jn.00527.2005. Intercepting an object requires a precise estimate of its time of arrival at the interception point (time to contact or "TTC"). It has been proposed that knowledge about gravitational acceleration can be combined with first-order, visual-field information to provide a better estimate of TTC when catching falling objects. In this experiment, we investigated the relative role of visual and nonvisual information on motor-response timing in an interceptive task. Subjects were immersed in a stereoscopic virtual environment and asked to intercept with a virtual racket a ball falling from above or rising from below. The ball moved with different initial velocities and could accelerate, decelerate, or move at a constant speed. Depending on the direction of motion, the acceleration or deceleration of the ball could therefore be congruent or not with the acceleration that would be expected due to the force of gravity acting on the ball. Although the best success rate was observed for balls moving at a constant velocity, we systematically found a cross-effect of ball direction and acceleration on success rate and response timing. Racket motion was triggered on average 25 ms earlier when the ball fell from above than when it rose from below, whatever the ball's true acceleration. As visual-flow information was the same in both cases, this shift indicates an influence of the ball's direction relative to gravity on response timing, consistent with the anticipation of the effects of gravity on the flight of the ball. I N T R O D U C T I O NSuccess in intercepting a moving target depends in part on the accuracy with which we estimate the time that remains before the object arrives at the point of interception (time to contact or TTC). A number of visual cues can contribute to TTC estimates, such as dilation of the retinal image, changes in binocular disparity, reduction of the optical gap between the target and interception point, and the evolution of pursuit eye movements (Bootsma and Oudejans 1993;Gray and Regan 1998;Rushton and Wann 1999;Savelsbergh et al. 1993). While it is known that the processing of visual information can provide accurate estimates of an object's distance and velocity, human observers are generally poor at measuring or even detecting acceleration (Werkhoven et al. 1992). This paucity of acceleration and higher-order cues means that available visual information is ambiguous; for a given measurement of position and velocity, there are an infinite number of possible true TTC values that could occur, depending on future changes in velocity. A reasonable way to resolve this ambiguity would be to assume no acceleration. Thus the main hypothesized mechanisms for TTC estimates based on visual information involve first-order approximation...

Several transcranial magnetic stimulation (TMS) studies have reported facilitation of the primary motor cortex (M1) during the mere observation of actions. This facilitation was shown to be highly congruent, in terms of somatotopy, with the observed action, even at the level of single muscles. With the present study, we investigated whether this muscle-specific facilitation of the observer's motor system reflects the degree of muscular force that is exerted in an observed action. Two separate TMS experiments are reported in which corticospinal excitability was measured in the hand area of M1 while subjects observed the lifting of objects of different weights. The type of action 'grasping-and-lifting-the-object' was always identical, but the grip force varied according to the object's weight. In accordance to previous findings, excitability of M1 was shown to modulate in a muscle-specific way, such that only the cortical representation areas in M1 that control the specific muscles used in the observed lifting action became increasingly facilitated. Moreover, muscle-specific M1 facilitation was shown to modulate to the force requirements of the observed actions, such that M1 excitability was considerably higher when observing heavy object lifting compared with light object lifting. Overall, these results indicate that different levels of observed grip force are mirrored onto the observer's motor system in a highly muscle-specific manner. The measured force-dependent modulations of corticospinal excitability in M1 are hypothesized to be functionally relevant for scaling the observed grip force in the observer's own motor system. In turn, this mechanism may contribute, at least partly, to the observer's ability to infer the weight of the lifted object.