Driving a real wheelchair by means of a braincomputer interface (BCI) system must be a reliable option for locked-in patients. Such navigation should also be autonomous, i.e., not depending on a ground chart. In this work we test the feasibility of driving a customized robotic wheelchair with a BCI system that our group has used in previous studies with virtual and real mobile robots. The results obtained from a sample of three healthy naïve participants suggest that it is an effective option, which could ultimately provide locked-in patients with greater autonomy and quality of life.

No abstract

Due to counting on only one locked-in patient, the current work constitutes a feasibility study. The actual usability of systems such as the one proposed in this paper should be determined by means of studies with a greater number of end-users in real-life conditions.

Two experiments demonstrated renewal effects in interference between outcomes in human participants. Experiment 1 revealed a XYX renewal effect, whereas Experiment 2 showed a XYZ renewal effect. The results from both experiments conformed to Bouton's (1993) theory of interference and recovery from interference, and contradicted the predictions derived from alternative accounts. Unlike previous demonstration of renewal effects, a cued response reaction time (RT) task was used, able to detect the effects of fast retrieval processes based on associative activation and that allowed little impact of inferential reasoning.

The neural response to positive and negative feedback differs in their event-related potentials. Most often this difference is interpreted as the result of a negative voltage deflection after negative feedback. This deflection has been referred to as the feedback-related negativity component. The reinforcement learning model of the feedback-related negativity establishes that this component reflects an error monitoring process aimed to increase behavior adjustment progressively. However, a recent proposal suggests that the difference observed is actually due to a positivity reflecting the rewarding value of positive feedbacks - that is, the reward positivity component (RewP). From this it follows that RewP could be found even in the absence of any action-monitoring processes. We tested this prediction by means of an experiment in which visual target stimuli were intermixed with nontarget stimuli. Three types of targets signaled money gains, money losses, or the absence of either money gain or money loss, respectively. No motor response was required. Event-related potential analyses showed a central positivity in a 270-370 ms time window that was elicited by target stimuli signaling money gains, as compared with both stimuli signaling losses and no-gain/no-loss neutral stimuli. This is the first evidence to show that RewP is obtained when stimuli with rewarding values are passively perceived.