Empowering the BCI User

One significant, yet largely untapped, way to empower the user of a BCI device is to work with the power of the mind. The user’s mind or mental state is arguable the key component of device control. If users are unable to encode commands in their EEG patterns, no signal processing or machine learning algorithm would be able to decode them (Lotte, Jeunet, Mladenovic et al., 2018). Indeed, BCI software should be designed with deep understanding of the end-user, and end-users should be involved in every step of the design (Schreuder, Riccio, Risetti et al., 2013).

The user’s mental state is a potentially powerful resource for BCI control. But this can only be taken advantage of fully if the user’s mind is acknowledged, understood, and then defined.

First, the user’s state of mind varies greatly. There are of course ways to stabilize, or make consistent, the user’s state of mind. However, there is also a great deal of unavoidable variability. This includes life event context such as a job change or death in the family, measurement environment/situation changes (social, home, work, lab…), and BCI operation itself (degree of success, anxiety, frustration, relief, confidence, motivation…).

I argue any motor control, including BCI control, always occurs within a (larger) mind context. This context changes the motor control signal itself. After all, a person’s perception, cognition, emotion etc. continually affects his or her movement. An anxious, depressed, and unsure person moves differently than that same person who feels confident, calm, and motivated. The user — who is also a person — generates an efferent motor control signal greatly affected by their overall state of mind. A motor command signal is only one part of his or her efferent signal.

Given the importance of the user’s mind in BCI control, how might this empower the user? One way is to give the user a large say in what state of mind “targets” to “shoot for.” A user could define what mental states they wish to target, and subsequently hit. He or she could define optimal target states of mind prior to use. This might involve self-reflection, consultation with family or co-workers, and research. These “mind targets” could be personalized to fit the user’s lifestyle, occupation, social habits, common activities, and personal goals.

In terms of improving BCI function, the user could define mind targets that increase motivation and interest, and are natural and easy to attain. This would enable the targets to be hit more strongly and reliably.

Self-chosen states of mind make the easiest targets to hit. Why? Because they are defined by someone who knows the user best — the user herself! For example, if the user is a Zen practitioner, “being mindful and focused on the task, moment-by-moment, during (BCI) movement” could be an appealing target. If the user was an intellectual by trade or hobby, “moving while thinking and staying calm” might be a desirable target.

Personalized, user-defined mind/brain targets could be bolstered by research and consultation with experts. For example, the user might learn it’s difficult to maintain a constant level of excitement, and dopamine-based brain signal, when one is alternately “failing” and “succeeding” at a task (such as device control). Social situations would magnify this emotional rollercoaster. To allow for this emotional noise, the user might define a target that includes “emotional variability.”

On the other hand, the user might be a devoted Buddhist who is (or strives to be) unattached to results. In this case, she could define her emotional target as “an even level of excitement while moving.”

Overall, optimal mind/brain targets can be researched, considered, discussed, and defined by the user and her clinical and professional team. Targets could be personalized to account for occupation, lifestyle, and other variables. They could also be activated flexibly according to the demands of the situation. The user would be able to hit these targets with greater strength, consistency, and selectivity. This is because (1) they are easier to hit, (2) they are motivated to achieve these states of mind, and (3) they have a clear idea of what the targets are in the first place!


Lotte, F., Jeunet, C., Mladenovic, J., N’Kaoua, B., Pilette, L. (2018). A BCI Challenge for the signal processing community: considering the user in the loop. Signal Processing and Machine Learning for Brain-Machine Interfaces, IET 1.

Schreuder, M., Riccio, A., Risetti, M., Dahne, S., Ramsay, A., Williamson, J., Mattia, D., Tangermann, M. (2013). User-centered design in brain-computer interfaces — a case study. Artificial Intelligence in Medicine, 59, 71-80.

Comments are closed.