Chaotic Visual Cryptosystem Using Empirical Mode Decomposition Algorithm for Clinical EEG Signals

• Published in: Journal of medical systems 2016-03, Vol.40 (3), p.52-10, Article 52
• Main Author: Lin, Chin-Feng
• Format: Article
• Language: English
• Subjects: Algorithms; Chaos theory; Computer Security; Electroencephalography; Electroencephalography - methods; Encryption; Health Informatics; Health Sciences; Humans; Mathematical analysis; Medicine; Medicine & Public Health; Nonlinear Dynamics; Signal processing; Signal Processing, Computer-Assisted; Statistics for Life Sciences; Systems-Level Quality Improvement; Two dimensional; Vectors (mathematics); Visual
• ISSN: 0148-5598; 1573-689X
• DOI: 10.1007/s10916-015-0414-0

This paper, proposes a chaotic visual cryptosystem using an empirical mode decomposition (EMD) algorithm for clinical electroencephalography (EEG) signals. The basic design concept is to integrate two-dimensional (2D) chaos-based encryption scramblers, the EMD algorithm, and a 2D block interleaver method to achieve a robust and unpredictable visual encryption mechanism. Energy-intrinsic mode function (IMF) distribution features of the clinical EEG signal are developed for chaotic encryption parameters. The maximum and second maximum energy ratios of the IMFs of a clinical EEG signal to its refereed total energy are used for the starting points of chaotic logistic map types of encrypted chaotic signals in the x and y vectors, respectively. The minimum and second minimum energy ratios of the IMFs of a clinical EEG signal to its refereed total energy are used for the security level parameters of chaotic logistic map types of encrypted chaotic signals in the x and y vectors, respectively. Three EEG database, and seventeen clinical EEG signals were tested, and the average r and mse values are 0.0201 and 4.2626 × 10 − 29 , respectively, for the original and chaotically-encrypted through EMD clinical EEG signals. The chaotically-encrypted signal cannot be recovered if there is an error in the input parameters, for example, an initial point error of 0.000001 %. The encryption effects of the proposed chaotic EMD visual encryption mechanism are excellent.