Motor Imagery BCI Demo with NeuraDock

Overview

This project demonstrates a motor imagery EEG workflow using the NeuraDock EEG Workstation, a 7-channel dry-electrode EEG platform for researchers, developers, and neurotechnology teams.

The demo asks a simple question:

Can imagined movement be detected before an actual grasp happens?

In this experiment, we use a ball-catching preparation task to guide motor imagery. The participant watches a ball falling toward a hand. During the task, EEG activity is monitored in real time, and the system attempts to detect motor intention before the grasp action is triggered

In 10 trials, the system successfully detected motor intention in 8 trials, reaching an 80% success rate in this demo.

Why Motor Imagery?

Motor imagery is a classic paradigm in brain-computer interface research. Instead of relying only on physical movement, the system looks for neural activity related to movement preparation or imagined movement.

This is especially relevant for rehabilitation-oriented BCI workflows, where the goal is not only to classify a signal, but also to build a closed-loop interaction between brain activity, feedback, and motor training.

In this demo, we use a ball-catching task because it is intuitive. The falling ball naturally guides the user into a preparation state before the grasp action.

Paradigm Design

The task is divided into three phases.

0–2 seconds: Baseline phase

During the first two seconds, the participant simply watches the ball falling smoothly. No action is required.

This phase provides a short baseline reference signal before the core motor preparation phase begins.

2–4 seconds: Motor preparation phase

As the ball approaches the hand, it begins to flicker.

This flickering cue is an important part of the paradigm. It uses SSVEP-based visual stimulation to strengthen task-related neural activity and make the motor imagery task more structured.

During this phase, the system monitors EEG activity and looks for signs of motor intention.

After 4 seconds: Grasp feedback

If the algorithm detects a clear motor intention, the system triggers the grasp action shown in the demo.

This creates a closed-loop workflow:

visual task → motor imagery → EEG detection → grasp feedback

Signal Processing Logic

The algorithm makes its judgment in two main steps.

1. Alpha-band power changes

First, the system looks at changes in alpha-band power.

Alpha activity is often associated with cortical rest or inhibition. When a brain region becomes more actively engaged, alpha-band power may decrease.

In this demo, the system monitors alpha-band activity related to the visual-motor preparation process.

2. Inter-channel power correlation

The second step is to calculate power correlations between specific electrode pairs.

For example, a correlation between visual-area and motor-preparation-related electrodes can indicate that the visual-to-motor pathway is becoming more active during the task.

Instead of only looking at single-channel power changes, this approach also considers how activity patterns between channels change together.

Why Welch’s Method?

Because this is a real-time task, the system cannot wait for a long recording window.

In this demo, the algorithm only has a short time window to make a decision. Welch’s method is used for power spectral estimation because it can split a short signal segment into smaller overlapping segments and average them.

This helps reduce variance caused by noise and provides a more stable spectral estimate than a single-window estimate.

For real-time BCI workflows, this stability is important because the system needs to make decisions quickly and reliably.

Demo Result

In this demo:

• Task type: Ball-catching motor imagery paradigm
• EEG hardware: NeuraDock EEG Workstation
• Setup: 7-channel dry-electrode EEG
• Trials: 10
• Successful detections: 8
• Success rate: 80%
• Main features: Alpha-band power and inter-channel power correlations
• Feedback: Grasp action after motor intention detection

This result shows a proof-of-concept workflow for real-time motor intention detection using dry-electrode EEG.

It is not presented as a clinical system, but as an experimental BCI workflow showing how NeuraDock can support real-time EEG-based interaction design and rehabilitation-oriented prototyping.

What This Demo Shows

This project shows that NeuraDock can support more than passive EEG recording.

It can be used to build structured BCI workflows that combine:

• Real-time EEG acquisition
• Motor imagery task design
• Visual stimulation
• Signal processing
• Intention detection
• Closed-loop feedback

The key point is not a single algorithmic trick. The value comes from combining hardware, software, task design, and signal analysis into a complete workflow.

Why This Matters

For researchers and developers, BCI experiments often require much more than an EEG headset. A useful workflow needs signal acquisition, task timing, feature extraction, decision logic, and feedback.

NeuraDock is being developed as an EEG workstation for this type of practical experimentation.

We are continuing to prepare more example workflows, including SSVEP, visual reconstruction, signal quality checking, marker workflows, and AI-assisted EEG analysis.

Links

NeuraDock is now on Crowd Supply prelaunch. Follow the campaign to get launch updates and support a more accessible EEG workstation for researchers, developers, and the open neurotechnology community.

Crowd Supply campaign:https://www.crowdsupply.com/neuradock/neuradock-eeg-workstation

Website:https://www.neuradock.com

GitHub:https://github.com/Neuradock

YouTube:https://www.youtube.com/@NeuraDo