Non-immersive Versus Immersive Extended Reality for Motor Imagery Neurofeedback Within a Brain-Computer Interfaces

Abstract

A sensory feedback was employed for the present work to remap brain signals into sensory information. In particular, sensorimotor rhythms associated with motor imagery were measured as a mean to interact with an extended reality (XR) environment. The aim for such a neurofeedback was to let the user become aware of his/her ability to imagine a movement. A brain-computer interface based on motor imagery was thus implemented by using a consumer-grade electroencephalograph and by taking into account wearable and portable feedback actuators. Visual and vibrotactile sensory feedback modalities were used simultaneously to provide an engaging multimodal feedback in XR. Both a non-immersive and an immersive version of the system were considered and compared. Preliminary validation was carried out with four healthy subjects participating in a total of four sessions on different days. Experiments were conducted according to a wide-spread synchronous paradigm in which an application provides the timing for the motor imagery tasks. Performance was compared in terms of classification accuracy. Overall, subjects preferred the immersive neurofeedback because it allowed higher concentration during experiments, but there was not enough evidence to prove its actual effectiveness and mean classification accuracy resulted about 65%. Meanwhile, classification accuracy resulted higher with the non-immersive neurofeedback, notably it reached about 75%. Future experiments could extend this comparison to more subjects and more sessions, due to the relevance of possible applications in rehabilitation. Moreover, the immersive XR implementation could be improved to provide a greater sense of embodiment.

How different immersive environments affect intracortical brain computer interfaces

“…While there were significant differences in performance in the different conditions, the magnitude of the differences was relatively small and both participants were able to complete the tasks satisfactorily in both conditions. This result adds an interesting piece of information to previous work that showed conflicting results as to how a VR environment could influence BCI control [20][21][22]. It may be that the importance of immersiveness is small or that the importance of experience in an environment is large, but in either case a device for long-term use should be able to utilize either type of environment.…”

“…These findings highlight the importance of clear and accurate visual feedback. Researchers have begun testing how different setups, such as a VR headset or a standard monitor display, effect BCI usability; however, results have varied from improvements when working in VR [20], to mixed effects [21], or no difference between the two conditions [22]. These studies have primarily relied on non-invasive or low-dimensional BCIs which can work with an abstracted strategy to control movement; little is known about how the immersiveness of the environment will affect more intuitively controlled higher degree of freedom BCIs.…”

Transforming metaverse marketing into strategic agility in SMEs through mediating roles of IMT and CI: theoretical framework and research propositions

“…This is based on the assumption that official IMT uses for SA are not firmly established. According to Arpaia et al (2022), when used prudently, AR, VR and MR technologies may attract company customers to brand communities in the metaverse, resulting in a higher long-term return on investment (ROI) and SA. These tools are leading the way in customer experience management.…”

Visual and haptic feedback in detecting motor imagery within a wearable brain–computer interface

“…When designing the experiments, the number of subjects was chosen according to the expected effect size [48,49] due to neurofeedback. Notably, a preliminary study [50] suggested a Cohen's d index greater than 0.8, namely a large effect size [51]. Consequently, a sample size of six was enough to achieve a statistical power around 95 % [52].…”