WU-NEAT

WU-NEAT analyzes neonatal electroencephalogram (EEG) recordings to compute amplitude-integrated EEG (aEEG) and spectral edge frequency (SEF) metrics for standardized assessment of brain function in preterm and term infants.

Key Features:

• Clinically validated algorithms: Includes amplitude-integrated EEG (aEEG), which filters, rectifies, time-compresses and displays data on a semi-logarithmic scale, and spectral edge frequency (SEF), which computes the frequency below which a specified proportion of EEG spectral power resides.
• Device-independence: Processes single-channel EEG input irrespective of recording source or sampling rate.
• Reliability and validation: aEEG evaluation demonstrated intra-rater reliability of 100% and inter-rater reliability of 98%, and SEF validation yielded a mean Pearson correlation coefficient of 0.96 ± 0.04 versus reference calculations.
• Computational environment: Implemented to run within MATLAB with support for GNU Octave.

Scientific Applications:

• Neurological monitoring: Quantifies EEG features used to monitor neonatal neurological development and detect abnormalities.
• Longitudinal studies: Provides consistent quantitative metrics to support repeated measures over time in preterm and term infant cohorts.
• Comparative research: Enables standardized comparison of EEG-derived metrics across datasets or institutions.

Methodology:

The aEEG algorithm processes raw EEG by filtering, rectifying, time-compressing and producing a semi-logarithmic display, and the SEF algorithm calculates the critical frequency below which a specified proportion of EEG spectral power is located.