Neonatal EEG Analysis Toolbox (NEAT) Code and Data

Many qualitative and quantitative metrics have been developed to analyze the output of limited-channel EEG recordings in neonates. A significant roadblock in the widespread dissemination of these techniques is the device-specific, “black-box” nature of the analytic algorithms, preventing comparison between studies and the obsolescence of old recordings as software is abandoned. We have developed NEAT (Neonatal EEG Analysis Toolbox) as an open-source software platform to overcome these challenges. NEAT runs within MATLAB or GNU Octave and is designed to work with any single-channel EEG input, regardless of source or sampling rate. The experimental GUI is only compatible with MATLAB. This package includes source code for MATLAB and GNU Octave and a sample recording file for testing purposes. System Requirements: MATLAB r2016a or later with the Signal Processing Toolbox or GNU Octave version 4.4.1 or later with the 'signal' package Release Notes: Release 2 - 7/16/2020 -Added GNU Octave variation of aEEG and SEF tools -Changed bandpass filter in SEF algorithm from Butterworth to FIR design for better GNU Octave compatibility Release 1 - 06/21/2017 -Includes two command-line MATLAB functions to generate aEEG signal and calculate spectral edge frequency -Includes experimental GUI, allowing for linked navigation of raw/aEEG signals

目前已开发出多种定性与定量指标，用于分析新生儿有限通道脑电图（electroencephalogram，EEG）的记录输出。此类技术在广泛推广过程中面临的显著障碍：分析算法具备设备特异性的“黑箱”属性，这不仅阻碍了不同研究间的结果对比，且随着软件被弃用，旧有脑电图记录也将失效。为此，我们开发了NEAT（新生儿脑电图分析工具箱，Neonatal EEG Analysis Toolbox）作为开源软件平台，以解决上述挑战。NEAT可在MATLAB或GNU Octave环境下运行，支持任意单通道EEG输入，不受信号源与采样率限制。其实验性图形用户界面（Graphical User Interface，GUI）仅兼容MATLAB。本软件包包含适用于MATLAB与GNU Octave的源代码，以及一份用于测试的示例记录文件。系统需求：搭载信号处理工具箱（Signal Processing Toolbox）的MATLAB r2016a及以上版本；或搭载'signal'包的GNU Octave 4.4.1及以上版本。版本更新说明：版本2 - 2020年7月16日 - 新增适配GNU Octave的振幅整合脑电图（amplitude-integrated electroencephalogram，aEEG）与频谱边缘频率（spectral edge frequency，SEF）工具；将SEF算法中的带通滤波器从巴特沃斯（Butterworth）设计更改为有限长单位冲激响应（Finite Impulse Response，FIR）滤波器，以提升与GNU Octave的兼容性。版本1 - 2017年6月21日 - 包含两个MATLAB命令行函数，分别用于生成aEEG信号与计算SEF；搭载实验性GUI，可实现原始EEG与aEEG信号的联动导航。