Abstract
This study reports the potential and challenges of educational neuroscience applications in mathematics education. By analyzing cognitive processes such as spatial reasoning and problem solving through the analogy of solving a Rubik's Cube, we elucidate their underlying neural mechanisms. Traditional mathematics education research relies primarily on observable behaviors and outcome analyses, limiting insights into students' real-time and implicit cognitive processes. In contrast, neuroscience techniques (e.g., fMRI, EEG, fNIRS) offer deeper exploration into the neural foundations of mathematical learning, overcoming traditional methodological limitations. Internationally, special journal issues and book chapters actively discuss interdisciplinary frameworks including neuro-understanding, neuro-prediction, and neuro-intervention, reflecting growing trends toward integrating neuroscience into mathematics education. This article further proposes the application of passive Brain-Computer Interface (passive BCI) technology combined with artificial intelligence for real-time analysis of neural signals. This approach aims to diagnose diverse cognitive strategies and visualization processes employed by students during mathematical problem-solving, thereby enhancing teachers' understanding of students' cognitive processes. Ultimately, this paper seeks to foster innovative developments in Taiwanese educational neuroscience research and provide practical insights for individualized mathematics instruction in classroom contexts.