Optimal Feature Analysis for Identification Based on Intracranial Brain Signals with Machine Learning Algorithms

Biometrics, e.g., fingerprints, the iris, and face, have been widely used to authenticate individuals. However, most biometrics are not cancellable, i.e., once these traditional cloned or stolen, they cannot be replaced easily. Unlike biometrics, brain extremely difficult clone forge due natural randomness across different individuals, which makes them an ideal option for identity authentication. Most existing based on electroencephalogram (EEG), typically demonstrates unstable performance low signal-to-noise ratio (SNR). Thus, in this paper, we propose use of intracortical signals, higher resolution SNR, realize construction a high-performance biometric. Significantly, is first study investigate features signals identification. Specifically, several local field potential computed identification, their compared with machine learning algorithms. The results show that frequency domain time-frequency excellent intra-day inter-day Furthermore, energy perform best among all 98% 93% identification accuracy, great intracraial biometrics. This paper may serve as guidance future intracranial researches development more reliable