QStates is a powerful cognitive state classification tool with a Partial Least-Squares (PLS) core algorithm for rapid and intuitive subject-specific or normative model training.
QStates extracts approximately 10,000 features from EEG, ECG, EMG, and EOG signals, and rapidly sets weights on the most salient features. The PLS algorithm is fast and trains models using as little as 1 minute of data for each state. The fast training allows subject specific daily calibration reducing performance variability.
Seamless integration with QStreamer allows real-time classification of cognitive load, with graphical intuitive visual feedback. Post-hoc analysis also displays quality of data metrics, and provides results and summaries in comma-separated value (CSV) formatted files.
QStates reliably classifies cognitive workload, engagement, and fatigue, and has been validated on numerous tasks in multiple environments to produce >90% classification accuracy.
QStates Specifications
-
• Up to 3 cognitive models at a time
• Inputs EEG, ECG, EMG, and EOG data
• Outputs linear and non-linear cognitive gauges
• Monitors Quality of Data
• Real-time interface with QStreamer
• Exports data to comma-separated-value (CSV) ASCII files
This technology has been licensed to Wearable
Sensing for further development and commercialization.
Please check out their website for new products based on this technology and contact
them directly for further information and pricing.