Asynchronous non-invasive brain-actuated control of an intelligent wheelchair

Abstract

Abstract-In this paper we present further results of our asynchronous and non-invasive BMI for the continuous control of an intelligent wheelchair. Three subjects participated in two experiments where they steered the wheelchair spontaneously, without any external cue. To do so the users learn to voluntary modulate EEG oscillatory rhythms by executing three mental tasks (i.e., mental imagery) that are associated to different steering commands. Importantly, we implement shared control techniques between the BMI and the intelligent wheelchair to assist the subject in the driving task. The results show that the three subjects could achieve a significant level of mental control, even if far from optimal, to drive an intelligent wheelchair.

The role of shared-control in BCI-based telepresence

“…In this paper we have described one of the first implementations of a shared-control BCI telepresence robot that is related to our previous work with wheelchairs [11], [12], [13], [14]. In the current paper, however, shared control only deals with lowlevel obstacle avoidance and it relies on very simple infrared sensors (as opposed to expensive laser range finders on the wheelchairs).…”

“…Thus, for them, a semi-autonomous approach is more suitable for a BCIcontrolled telepresence robot, where the intelligent system helps the human user to cope with problematic situations such as obstacle detection and avoidance [9]. Such a semiautonomous framework has been largely explored for assistive wheelchairs [4], [10], including brain-controlled wheelchairs [11], [12], [13], [14].…”

On the need for on-line learning in brain-computer interfaces

“…Indeed, so far control tasks based on human EEG have been limited to simple exercises such as moving a computer cursor or opening a hand orthosis [7]- [8]. However, recently we have shown that an EEG-based BCI can also make possible the continuous control of a mobile robot (emulating a motorized wheelchair) generating non-trivial trajectories among different rooms in a house-like environment [3]. The key idea here is that the user's mental states are associated with highlevel commands (e.g., "turn right at the next occasion") and that the robot executes these commands autonomously using the readings of its on-board sensors.…”

“…In this paper we use a simple statistical classifier to illustrate the benefits of on-line learning with three beginner subjects. After a few sessions, two of these subjects reached a performance that, according to our previous experience, is close to allow them to start operating a brain-actuated device [1]- [3], and this despite the subjects have been trained without feedback (see Section 3). To the best of our knowledge, the issue of on-line learning in BCI has not been explored previously.…”

“…Over the last years we have developed a portable brain-computer interface (BCI) that analyzes on-line electroencephalogram (EEG) signals-i.e., the brain electrical activity recorded from electrodes placed onto the scalp-in order to determine the subject's mental state that is then mapped into actions such as selecting a letter from a virtual keyboard or driving a mobile robot [1]- [3]. Ours is one of the few BCIs that have been demonstrated in practice enabling subjects to control different devices such as a computer cursor, a virtual keyboard and even robots [4]- [8].…”