Halyzia
Halyzia™
is a user-friendly solution that provides quantitative analyses of iEEG signals, automatically detecting and quantifying features that sometimes cannot be seen by the naked eye. Halyzia™ can process EEG recorded on either macrocontacts or microwires.
Our algorithm includes data processing (time series filtering, time-frequency maps display) and artificial intelligence as documented in our patent.
Halyzia™ ’s features include:
Management of raw data: our AI can handle any data, even if artifacted.
Results reviewing: the user can easily review, validate or discard detected events or tag new ones .
Automated reporting: a topographic map summarizes live detected events in an intuitive graphical manner (see below).
Quantitative results: users can generate an instantaneous summary of detected events, e.g. averaged frequency peaks, fast ripple index, entropy, event duration etc.
Annotation: users can annotate individual signals with custom labels and export them in standard formats.
Standards: collected and generated data is stored using the BIDS format .
Halyzia™ has been validated using retrospectively data from more than 50 patients with drug-resistant epilepsy who were/are candidates for therapeutic surgery. It is currently being evaluated in a prospective multi-centric clinical trial.
Simple & Intuitive Display
A topographic representation allows the quick identification from which
electrodes and channels events have been detected.
Halyzia™ has been trained on a database built by epileptologists and epilepsy researchers over several years. Our database is continuously being enriched as new patient EEG data is processed, allowing us to continuously improve our algorithms. Our database is available for licencing.
Process EEG at any Scale
In the example below, Halyzia has been used to detect fast-ripples recorded on tetrodes.