Apple FaceID co‑inventor builds AI model to decode brain signals

Former Apple engineer and FaceID co‑inventor John Doe is spearheading a groundbreaking artificial‑intelligence project that aims to translate raw electrical brain signals into actionable diagnostic information.

From FaceID to Brain Mapping

After leaving Apple, Doe joined a neurotechnology startup where he is applying his expertise in signal processing to decode the brain's electrical activity captured by non‑invasive electrodes.

The AI model, dubbed NeuroLens, leverages deep learning architectures originally designed for facial recognition, repurposing them to identify patterns in neural data that correspond to specific cognitive states or pathologies.

How the Model Works

NeuroLens ingests high‑resolution electroencephalogram (EEG) recordings and trains on labeled datasets that include both healthy participants and patients with known neurological conditions. By learning subtle variations in waveform morphology, the system can flag anomalies that may indicate early stages of disorders such as epilepsy or Alzheimer’s disease.

• Collect EEG data from volunteers
• Preprocess signals to remove noise
• Train deep neural networks on labeled examples
• Validate predictions against clinical diagnoses

Potential Clinical Impact

If successful, the technology could provide clinicians with a rapid, non‑invasive screening tool, reducing reliance on costly imaging methods and enabling earlier intervention.

Doe emphasizes that the model is intended to augment, not replace, expert medical judgment, serving as a decision‑support system that highlights areas for further investigation.

We’re turning the same pattern‑recognition engines that unlocked FaceID into a window on the brain.

The project is still in its research phase, with pilot studies underway at several academic medical centers. Results are expected to be published later this year.

Read the report