Abstract
Arm movements towards visual reaching targets provides a rich source of cues about final action goals and intentions. In the present research, we investigated how early during action execution movement kinematics can predict static and shifted visual target positions. Visual targets were distributed at different directions and depths of the visual space and we assessed whether the visual target representation shares the same temporal structure in the two spatial dimensions along the movement execution. Specifically, we analysed the arm trajectories while 13 participants were executing a visually guided reaching towards targets located at different depths and directions with respect to the body in peripheral viewing conditions. The visual targets could remain static during the entire duration of the movement or could shift along the horizontal (direction) or the sagittal dimension (depth) once the movement was started. In order to predict the position of static and shifted visual targets, we discretised the continuous horizontal and vertical component of movement trajectories in spatial intervals and we used a Linear Discriminant Analysis (LDA) classifier to extract classification accuracies describing how early it is possible to make reliable estimates about visual target positions. The results show that the horizontal trajectory component (of both pointing finger and wrist) was predictive only for the direction of static and shifted visual targets. In contrast, the vertical trajectory component of pointing finger and wrist was predictive only for the depth of static and shifted visual targets. In both cases, wrist trajectories showed higher classification accuracy with respect to the pointing finger with increasing accuracy of all visual target positions starting from the 50% of movement. These findings suggest that it is reliably possible to extract visual target information from movement kinematics well before the conclusion of action without significant difference in temporal structure between direction and depth dimension.