BrainPrint: EEG biometric identification based on analyzing brain connectivity graphs

作者：

Highlights：

• The topological features of brain connectivity graphs can be effectively used for EEG biometric identification.

• Seven connectivity metrics including a new one defined on the algorithmic complexity of signals, and twelve graph features are evaluated for network estimation and feature extraction.

• The study also analyzes the impact of EEG frequency bands and regions on biometric recognition performance and discusses the intra-subject variation issue of EEG biometrics.

摘要

•The topological features of brain connectivity graphs can be effectively used for EEG biometric identification.•Seven connectivity metrics including a new one defined on the algorithmic complexity of signals, and twelve graph features are evaluated for network estimation and feature extraction.•The study also analyzes the impact of EEG frequency bands and regions on biometric recognition performance and discusses the intra-subject variation issue of EEG biometrics.

论文关键词：EEG biometrics,Brain functional connectivity,Person identification

论文评审过程：Received 15 May 2019, Revised 9 March 2020, Accepted 14 April 2020, Available online 26 April 2020, Version of Record 4 May 2020.

论文官网地址：https://doi.org/10.1016/j.patcog.2020.107381