Many psychiatric and neurodevelopmental assessments still rely mostly on interviews and rating scales. We are researching whether EEG can add an objective signal to that process, and we publish what we find, including what does not work yet.
Leave-one-subject-out validation and audited pipelines. When we found leakage in our own results, we corrected the numbers in public.
Every result is shown with the dataset, age range and method behind it. Nothing is presented as a diagnosis.
Parts of our tooling are open source, so others can check the methods and reuse them.
An EEG recording is uploaded to the platform as an EDF file.
Filtering, epoching and automatic cleaning of noisy segments with MNE-Python and AutoReject.
Spectral power, band ratios, coherence between channels and complexity measures.
A calibrated probability with the features that drove it (SHAP), for the clinician to review.
All modules are in research and development. None is approved for clinical diagnostic use.
EEG biomarker analysis to support ADHD assessment. The most advanced SynapCore module, with preliminary results in a paediatric cohort.
Explore NeuraxisPreliminary research results. Not available for clinical use.
Frontal alpha asymmetry and coherence measures to support assessment of major depressive disorder.
Gamma band and auditory ERP markers for schizophrenia spectrum evaluation.
Resting-state power spectral profiles to help tell bipolar states apart.
Theta oscillation and hyperarousal markers for PTSD decision support.
We would rather show modest, reproducible numbers than impressive ones that do not hold up.
We are looking for research and clinical partners to help validate SynapCore.
SynapCore turns EEG recordings into decision-support reports for neurological and psychiatric assessment. Built to support clinicians, not to replace them.
