Scholarlink

The development of voice assistants and other interfaces offers the potential to enhance productivity through multitasking. However, this increase in the sources and types of environmental sounds (e.g., music, speech, and natural sounds) to which people are exposed daily raises, concerns about potential adverse effects, such as discomfort, altered cognitive function, and cognitive conflict induction. While previous studies have examined cognitive conflict using physiological data (e.g., electroencephalogram), this study proposes Near-InfraRed HemoEncephaloGraphy (NIR-HEG) as a low-cost, movement-resistant alternative. NIR-HEG enables the indirect evaluation of brain activity without restricting user movement and providing valuable feedback on interfaces used in daily life. This study investigates the feasibility of implementing NIR-HEG to classify and analyze cognitive conflict. Ten participants who passed the Ishihara Color Vision test completed two Stroop tasks under six environmental sound conditions designed to induce cognitive interference. Task performance metrics and NIR-HEG data from two prefrontal areas (FP1 and FP2, International 10–20 System) were used as input for a Support Vector Machine (SVM) classification model. While the classification performance for environmental sound conditions was suboptimal, achieving a maximum accuracy of 64.37 \(\%\) the classification of task types showed improved performance with a maximum accuracy of 88.01 \(\%\) . These findings suggest task types may influence brain activity patterns associated with cognitive conflict. They underscore the utility of NIR-HEG in elucidating the neural correlates of cognitive conflict and highlight its potential practical applications in interface design and implementation of user interfaces for real-world environments.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Cognitive Conflict Classification Depending on Background Auditive Environment Using Support Vector Machine Based on Near-Infrared Hemoencephalography

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Kota Yokoyama,
  • Keiichi Zempo,
  • Sandra Puentes

摘要

The development of voice assistants and other interfaces offers the potential to enhance productivity through multitasking. However, this increase in the sources and types of environmental sounds (e.g., music, speech, and natural sounds) to which people are exposed daily raises, concerns about potential adverse effects, such as discomfort, altered cognitive function, and cognitive conflict induction. While previous studies have examined cognitive conflict using physiological data (e.g., electroencephalogram), this study proposes Near-InfraRed HemoEncephaloGraphy (NIR-HEG) as a low-cost, movement-resistant alternative. NIR-HEG enables the indirect evaluation of brain activity without restricting user movement and providing valuable feedback on interfaces used in daily life. This study investigates the feasibility of implementing NIR-HEG to classify and analyze cognitive conflict. Ten participants who passed the Ishihara Color Vision test completed two Stroop tasks under six environmental sound conditions designed to induce cognitive interference. Task performance metrics and NIR-HEG data from two prefrontal areas (FP1 and FP2, International 10–20 System) were used as input for a Support Vector Machine (SVM) classification model. While the classification performance for environmental sound conditions was suboptimal, achieving a maximum accuracy of 64.37 \(\%\) the classification of task types showed improved performance with a maximum accuracy of 88.01 \(\%\) . These findings suggest task types may influence brain activity patterns associated with cognitive conflict. They underscore the utility of NIR-HEG in elucidating the neural correlates of cognitive conflict and highlight its potential practical applications in interface design and implementation of user interfaces for real-world environments.