Objective. In this paper, a novel methodology for feature extraction to enhance classification accuracy of functional near-infrared spectroscopy (fNIRS)-based two-class and three-class brain–computer interface (BCI) is presented. Approach. Novel features are extracted using vector-based phase analysis method. Changes in oxygenated ##IMG## [http://ej.iop.org/images/1741-2552/17/5/056025/jneabb417ieqn1.gif] {$\left( {{{\Delta }}HbO} \right)$} and de-oxygenated ##IMG## [http://ej.iop.org/images/1741-2552/17/5/056025/jneabb417ieqn2.gif] {$({{\Delta }}HbR$} ) haemoglobin are used to calculate four novel features: change in cerebral blood volume ( ##IMG## [http://ej.iop.org/images/1741-2552/17/5/056025/jneabb417ieqn3.gif] {$\Delta CBV$} ), change in cerebral oxygen exchange ( ##IMG## [http://ej.iop.org/images/1741-2552/17/5/056025/jneabb417ieqn4.gif] {$\Delta COE$} ), vector magnitude (

中文翻译：

---

使用从基于矢量的相位分析中获得的特征提高 fNIRS-BCI 的分类精度

客观的。在本文中，提出了一种新的特征提取方法，以提高基于功能近红外光谱 (fNIRS) 的二类和三类脑机接口 (BCI) 的分类精度。方法。使用基于矢量的相位分析方法提取新特征。含氧量变化##IMG## [http://ej.iop.org/images/1741-2552/17/5/056025/jneabb417ieqn1.gif] {$\left( {{{\Delta }}HbO} \对）$} 和脱氧 ##IMG## [http://ej.iop.org/images/1741-2552/17/5/056025/jneabb417ieqn2.gif] {$({{\Delta }}HbR $} ）血红蛋白用于计算四个新特征：脑血容量变化（##IMG## [http://ej.iop.org/images/1741-2552/17/5/056025/jneabb417ieqn3.gif] {$\Delta CBV$}），脑氧交换变化（##IMG## [http://ej.iop.org/images/1741-2552/17/5/056025/jneabb417ieqn4.