Steady State Visual Evoked Potential (SSVEP)-Based Brain–Computer Interface as a Control Method for Exoskeletons: A Review

Introduction to Brain–Computer Interfaces

Brain–computer interfaces (BCIs) represent a transformative approach to assistive technology, enabling individuals with severe motor impairments to interact with external devices using brain signals alone—without requiring voluntary muscle activity. The fundamental principle underlying BCI technology is the detection and classification of specific patterns of brain activity, typically measured through electroencephalography (EEG), which are then translated into control commands for computers, robotic devices, or communication systems.

Among the various BCI paradigms that have been developed—including motor imagery (MI), P300 event-related potentials, and steady-state visual evoked potentials (SSVEPs)—SSVEP-based systems have attracted particular attention due to their inherent advantages. SSVEP-based BCIs offer higher information transfer rates compared to most other EEG-based BCI approaches, require little or no user training, and produce relatively stable and repeatable brain signals. These characteristics make SSVEP an appealing paradigm for practical applications in rehabilitation and assistive technology.

This review, published in MJPHM’s 2016 Special Volume 1, systematically examined the state of the art in SSVEP-based BCI systems specifically as applied to exoskeleton control. By bringing together evidence from engineering, neuroscience, and rehabilitation medicine, the review addressed both the technical performance of these systems and their potential for clinical translation.

How SSVEP-Based BCI Works

The SSVEP phenomenon occurs when a person visually fixates on a stimulus that flickers at a constant frequency. This repetitive visual stimulation produces a measurable oscillatory electrical response in the visual cortex of the brain, at the same frequency as the flickering stimulus and at its harmonic frequencies. For example, if a visual stimulus flickers at 10 Hz, the EEG signal recorded over the occipital region of the scalp will show strong components at 10 Hz, 20 Hz, and potentially 30 Hz.

In a practical SSVEP-BCI system, the user is presented with multiple visual targets, each flickering at a distinct frequency. By focusing their gaze on a particular target, the user generates a corresponding SSVEP response that can be detected from the EEG signal. A signal processing algorithm then determines which frequency is dominant, thereby identifying which target the user is attending to. This frequency identification is translated into a control command—for example, directing an exoskeleton to perform a specific movement pattern.

Application to Exoskeleton Control

Exoskeletons—wearable robotic devices that augment or replace the motor function of the limbs—represent one of the most promising applications of BCI technology. For individuals with paralysis or severe motor impairment due to stroke, spinal cord injury, or neurodegenerative disease, BCI-controlled exoskeletons offer a pathway to functional independence that conventional assistive devices cannot provide.

The review examined multiple approaches to integrating SSVEP-BCI with upper-limb and lower-limb exoskeletons. In upper-limb applications, users typically select from a menu of predefined movements (such as reaching, grasping, or lifting) by fixating on the corresponding SSVEP stimulus. The exoskeleton then executes the selected movement pattern, with the BCI effectively serving as a hands-free command interface. Lower-limb applications have explored SSVEP control for standing, sitting, and gait initiation commands, enabling basic mobility for wheelchair-bound individuals.

A particularly promising development noted in the review was the emergence of hybrid BCI systems that combine SSVEP with other BCI paradigms or sensor modalities. For instance, SSVEP can be used to select the desired movement type, while motor imagery provides continuous proportional control of movement speed or force. Such hybrid approaches address the limitations of any single BCI paradigm and may ultimately provide the level of control sophistication needed for practical daily use.

Technical Challenges and Limitations

Despite the promising performance demonstrated in laboratory settings, several technical challenges impede the widespread clinical adoption of SSVEP-based BCI exoskeletons. The review identified the following as the most significant barriers to translation.

First, visual fatigue and user discomfort associated with prolonged exposure to flickering visual stimuli remain a concern. While SSVEP stimuli at higher frequencies (above 30 Hz) can reduce perceived flickering, they also produce weaker cortical responses, creating a trade-off between user comfort and signal quality. The SSVEP amplitude response varies across frequency ranges: low-frequency stimuli (4–15 Hz) produce the strongest responses but cause the most discomfort, medium-frequency stimuli (15–30 Hz) offer a reasonable compromise, and high-frequency stimuli (30–60 Hz) are most comfortable but technically challenging to detect.

Second, the requirement for external visual stimulation equipment constrains the portability and usability of SSVEP-BCI systems. Recent developments in augmented reality displays for SSVEP stimulus presentation may address this limitation by integrating stimuli into the user’s natural visual environment through head-mounted devices.

Third, EEG signal quality can degrade significantly during actual exoskeleton use, due to movement artefacts, muscle electrical activity, and changes in electrode impedance. Robust signal processing algorithms that maintain accurate frequency detection under these challenging conditions are essential for clinical viability.

Public Health and Rehabilitation Implications

The development of accessible, affordable BCI-controlled exoskeletons carries substantial public health implications for Malaysia and the broader Southeast Asian region. Stroke is a leading cause of adult disability in Malaysia, and the country’s ageing population is projected to increase the prevalence of stroke-related disability in coming decades. Spinal cord injury, while less common, results in profound long-term disability with significant impacts on quality of life and economic productivity.

Current rehabilitation services in Malaysia, while improving, face challenges related to therapist availability, geographical access, and cost. BCI-controlled exoskeletons, if they can be developed to be sufficiently reliable, affordable, and easy to use, could extend the reach of rehabilitation services by enabling home-based training programs that supplement clinic-based therapy. The neurofeedback component of BCI-based rehabilitation—where the user’s brain activity directly drives movement—has also been shown to promote neural plasticity and motor recovery in ways that passive rehabilitation cannot achieve.

This review contributes to the knowledge base needed to support the development and eventual clinical deployment of BCI-exoskeleton systems in the Malaysian healthcare context, where the engineering expertise within institutions such as the University of Malaya positions the country well to participate in this rapidly evolving field.

Limitations of the Review

The review was limited by the relatively small number of studies that had, at the time of publication, specifically examined SSVEP-BCI control of exoskeletons, as distinct from SSVEP-BCI systems for other applications (such as speller interfaces or wheelchair control). The rapidly evolving nature of the field means that significant technical advances have occurred since the review was conducted, including improvements in signal processing algorithms, electrode technology, and exoskeleton hardware. Additionally, the review focused primarily on the engineering and performance aspects of SSVEP-BCI systems, with less emphasis on the user experience, acceptability, and health economic considerations that will ultimately determine clinical adoption.