Neurofeedback training scenarios
Neurofeedback training scenarios
A research project at the intersection of design and neurotechnology. The goal was to find a form of feedback that helps a person track changes in brain activity and regulate their state.
- Research center
- Center for Bioelectric Interfaces, HSE University
- Role
- research, design, prototyping
- research
- design
- prototyping
- Format
Context
Neurofeedback is a technology that gives a person real‑time feedback on brain activity and gradually teaches them to regulate it.
In this project, I used a two‑rhythm training protocol: the participant had to simultaneously decrease theta rhythm and increase beta rhythm. I developed two scenarios for this protocol.
A scenario is a feedback mechanic through which the user sees how they influence the rhythm metrics. This feedback needs to be clear and quick.
Research
Before developing the scenarios, I conducted 6 in‑depth interviews with participants from a previous neurofeedback experiment. I needed to understand how people experienced the training: what helped them and what did not.
From the interviews, I learned that the visual environment held participants’ attention well: they liked the color, levels, character, and atmosphere. They followed the development of the scenario, but did not always read the instant feedback that showed changes in the rhythms.
Because of this, participants found it difficult to understand exactly how they were influencing the result: they tried breathing, relaxation, and focusing on individual elements, but could not describe a stable control strategy.
Hypothesis
This led to the hypothesis: a non‑gamified scenario would help users read instant feedback better and regulate brain rhythms more effectively.
To test the hypothesis, I compared two scenarios for the same protocol – a gamified scenario and a non‑gamified scenario.
Gamified scenario
Story
The user controls a submarine that collects waste in the ocean. The deeper the submarine descends, the more dangerous the waste it collects and the more points the participant earns.
Feedback mechanic
The mechanic is built around the movement of the submarine. The submarine moves at a 45-degree angle. Beta rhythm controls the descent speed, while theta rhythm works as a limiter.
Submarine movement diagram in the gamified scenario
Rhythm balance – color
Instant feedback is built into the behavior of the submarine’s engine. The bubbles show the balance between beta and theta rhythms and the submarine’s descent speed.
Large white bubbles mean that the descent speed is high and the contributions of the rhythms are roughly equal. This is the state to aim for. Yellow bubbles show a greater contribution from theta rhythm, while purple bubbles show a greater contribution from beta rhythm.
Color feedback diagram: contribution of beta and theta rhythms
The game scenario created many additional stimuli: the submarine, waste, points, and surrounding objects. Because of this, some participants focused more on the game than on the feedback.
Non‑gamified scenario
In the second scenario, I removed the plot, character, and external reward.
Feedback mechanic
The scenario has two layers of feedback.
The first layer is the vertical filling of the shape. An increase in beta rhythm helps the shape fill. Theta rhythm is responsible for the stability of this state. If theta decreases, it becomes easier for the shape to hold its filling. If theta increases, “gravity” becomes stronger and the shape loses its filling faster.
Diagram of the shape filling in the non‑gamified scenario
Rhythm balance – density
The second layer is the density of points inside the shape. It shows the balance between beta and theta rhythms. When the contributions of beta and theta are close to each other, the pattern becomes denser. When one rhythm starts to dominate, the density of points decreases.
Point density diagram: balance of beta and theta rhythms
The task became easier to read: instead of following the plot, points, and objects on the screen, the participant only had to hold the state of the shape.
Experiment
The main comparison involved 20 men aged 18 to 35 with no psychiatric or psychological diagnoses. There were two groups: 10 participants completed the gamified scenario, and 10 completed the non‑gamified scenario.
The procedure was the same for everyone. Before training, a baseline EEG was recorded with eyes open. The participant then completed 20 minutes of training, divided into 4 blocks of 5 minutes. After that, a second measurement was taken. At the end, participants completed the NASA TLX questionnaire.
Results
The main metric was the Main Score. It reflects the relative change in the ratio of theta rhythm to beta rhythm after training compared with the baseline level. The higher the score, the better.
The non‑gamified scenario showed a higher Main Score: 25.1% compared with 15.7% for the gamified scenario. However, this difference was not statistically significant, so we cannot claim that one scenario outperformed the other.
Looking at individual changes, the two scenarios were similar: a positive Main Score was observed in 8 out of 10 participants in each group, and an increase in beta rhythm was observed in 5 out of 10 participants in each group. The most consistent effect was a decrease in theta rhythm: it occurred in 10 out of 10 participants in the non‑gamified group and in 9 out of 10 participants in the gamified group.
In terms of subjective workload, the non‑gamified scenario felt more demanding: 41.7 compared with 32.5.
Scenario development and value
The main criterion is the balance between physiological outcome and subjective workload. The non‑gamified scenario produced a higher Main Score, but this advantage was not statistically confirmed. At the same time, the gamified scenario was perceived as easier: subjective workload was lower.
The sample was small, so the result cannot be considered final. However, as a product prototype, it makes sense to continue developing the gamified scenario: it showed a comparable physiological outcome while being better suited for regular use because it places less strain on the user.
Theta rhythm decreased significantly in both groups. This was a general training effect, not an advantage of one scenario over the other. The feedback based on theta rhythm worked as an implicit penalty: when theta rhythm increased, movement toward the target was slowed down. It is possible that this hidden penalty helped reduce theta activity.