Near-infrared Spectrum Brain Pattern Recognition Based on One-dimensional Convolutional Neural Network

• Published in: Chinese Control Conference pp. 8218 - 8223
• Main Authors: Yan, Xinyue, Feng, Yiyu, Zhang, Xianfu
• Format: Conference Proceeding
• Language: English
• Published: Technical Committee on Control Theory, Chinese Association of Automation 28.07.2024
• Subjects: Biological system modeling; Brain modeling; Classification algorithms; Convolutional neural network; Deep learning; Functional networks; Graphical models; Identity recognition; Machine learning; Pattern recognition; Predictive models
• DOI: 10.23919/CCC63176.2024.10662103

Differences between individual dynamic functional networks have stable spatial distributions, and whether such differences can be used for identification has not been fully studied. In this paper based on a set of public near-infrared spectroscopy datasets, we preprocessed the data of 30 subjects, constructed brain networks under different time windows, and identified them as "brainprint" features. Next, we compared the impact of traditional classification algorithms and deep learning classification algorithms on "brainprint" recognition results. For the 30 classification predictions of different feature experimental groups, the one-dimensional convolutional neural network (1DCNN) model constructed in this paper can achieve an average recognition accuracy of 98.67 \%. This shows that the CNN model constructed in this study is feasible for the classification of near-infrared "brainprint" and provides a potential biometric technology for future identity recognition.