Independent neurotechnology coverage

From brain to machine

Read story Explore devices Start the BCI path

Interactive 3D model

A brain you can turn in your hands

Six regions, real anatomy, zero scan required. Spin it, tap a region, then step into the full explorer for every structure, its function, and the sensors that can read it.

Open the full brain explorer

Latest Neurotech

Explore devices →

NeurotechnologySep 4, 2026

Bimanual Motor-Imagery Brain-Computer Interface Boosts Robotic Rehab Task Success to 90%

Researchers have introduced a motor-imagery-based brain-computer interface (BCI) system that decodes intent for both hands together, rather than the conventional single-limb approach, paired with a shared-control strategy between brain signals and robotic intelligence. Trials showed that adding graduated levels of robotic assistance raised the success rate of a bimanual reaching task from 48.33% with no assistance to 90% at the highest assistance level.

واجهة الدماغ والحاسوبالتخيل الحركيإعادة التأهيل الروبوتي

Research Radar

4 indexed papers

Structured digests of high-impact neural engineering and BCI publications.

Neural DecodingresearcherNature Machine Intelligence (2023)

Decoding speech perception from non-invasive brain recordings

A. Défossez, C. Caucheteux, J. Rapin, O. Kabeli, J.R. King

Neural DecodingMEGEEGSpeech BCI

Significance score

Read paper ↗

1. Problem

Decoding perceived speech from non-invasive brain recordings (EEG/MEG) has remained limited in accuracy because of low signal-to-noise ratio and high inter-subject variability.

2. Method

A contrastive-learning model was trained to align self-supervised speech representations with MEG and EEG recordings from 175 healthy volunteers listening to stories and isolated sentences, combining four public datasets.

3. Result

The model reached up to 70.7% top-10 segment-identification accuracy with MEG (41.3% top-1), versus only 17.7–25.7% with EEG, with the best individual participants exceeding 80% top-1 accuracy.

4. Significance

Shows that deep language models can extract rich linguistic structure from non-invasive recordings without any surgical electrode implantation.

5. Limitations

EEG performance remains far below MEG, MEG hardware is large and non-portable, and the model needs long per-participant recording sessions.