top of page

“In my opinion, there is no aspect of reality beyond the reach of the human mind.



Giving a Voice to the Voiceless

Decoding brain signals in real time, MindSpeaker restores natural communication for patients who lost their ability to speak.
Brain graphic 1.png

MindSpeaker decodes both speech and tone of voice directly from imagined speech. Our Brain Computer Interface outperforms traditional communication devices in terms of words per minute,
authenticity and mobility.


MindSpeaker, a KU Leuven neuroscience laboratory spin-off, is setting the standard in Brain Computer Interfacing. Our software for decoding imagined speech ensures security, reliability, and privacy by design.

Brain graphic 2.png
Non Invasive

MindSpeaker's technology offers next-generation performance while minimizing patient discomfort. Our wearable headsets, speakers, and EEG signal decoders accurately predict signal quality for high-fidelity decoding of imagined speech.


MindSpeaker technology is affordable and compatible with your existing assistive devices. We use software innovation to restore natural communication, promoting patient dignity.

MindSpeaker restores natural speech communication for patients suffering from paralysis.


From mind to speech

How does speech work?

Normal speech

Brain graphic 3.png

Imagined speech

Thoughts that are not expressed aloud but kept internal.

Functional for patients suffering from a speech disorder.


Dysfunctional for patients suffering from a speech disorder.

Performed speech

6. Articulation. The speech sounds are produced and can now be heard by others

5. Phonetic encoding. The sounds are transformed into precise motor commands to control the movement of the articulators

after 600 ms

600 ms

0 ms

1. Lexical concept. A mental representation of a word’s meaning or concept in the mind of an individual.

200 ms

2. Lexical selection. Choosing the most appropriate word to express an idea or meaning

275 ms

3. Morphological encoding. Generating the appropriate grammatical form of a word

455 ms

4. Phonological encoding. Generating sounds that make up the word.

bottom of page