Abstract Neurofeedback is a training technique to modulate neural functions by training self-regulation of brain activity in real-time. This technique has been applied in neurorehabilitation in patients with neurological and cognitive disorders. In addition, P300 has been used as a tool to evaluate these patients' cognitive functions. One type of event-related potential called the P300 reflects cognitive abilities and attention toward stimuli. However, setting up an environment to measure these is challenging due to a lack of software design experience. In this study, an electroencephalography (EEG)-based Neurofeedback and P300 Toolbox (NPT) was developed by integrating the Neurofeedback Tool and the P300 Tool. Most existing EEG analysis tools are specialized for offline analysis and do not support real-time signal processing or feedback during experiments. Although some tools offer real-time capabilities, using them for neurofeedback often involves additional configuration or development. Therefore, the NPT is designed specifically for real-time analysis environments. It provides real-time functional connectivity and neurofeedback protocol features such as alpha asymmetry and high beta down, with adjustable parameters for visualization. Furthermore, the NPT offers an integrated structure that enables seamless transitions between real-time data acquisition and offline analysis. This feature enhances the repeatability of experiments and the consistency of analysis, which is particularly useful for neurorehabilitation. Traditional P300 tools primarily offer signal processing and analysis functions, while P300 measurement is often performed using separate systems or software, likely due to time delays that can affect P300 latency. NPT is designed to support time-delayed compensation algorithms to integrate measurement and analysis within a single environment. All functions and options were confirmed to work properly. Experiments were conducted with the Neurofeedback Tool and P300 Tool on six participants without diagnosed conditions using NPT, confirming that all signals were detected properly.
Lee et al. (Wed,) studied this question.