IET Signal Processing (May 2022)

Intelligent physiological signal infosecurity: Case study in photoplethysmography (PPG) signal

  • Chia‐Hung Lin,
  • Jian‐Xing Wu,
  • Neng‐Sheng Pai,
  • Pi‐Yun Chen,
  • Chien‐Ming Li,
  • Ching Chou Pai

DOI
https://doi.org/10.1049/sil2.12089
Journal volume & issue
Vol. 16, no. 3
pp. 267 – 280

Abstract

Read online

Abstract A wrist‐based photoplethysmography (PPG) tool offers a simple and non‐invasive approach for applications in vital sign monitoring and healthcare. However, in a telecare network, these physiological signals ensure the authorisation demands for the domain of medical applications. Hence, this study proposes an intelligent method for PPG signal encryption and decryption. This method combines chaotic map and radial basis function network (RBFN) into symmetric cryptography with an adaptive scheme. The sine‐power chaotic map is used as a key generator of 256 non‐ordered numbers (key space) to set the private cipher codes. Two RBFNs are used to train an encryptor and a decryptor with the authorised cipher codes. Its substitution‐based infosecurity scheme can change the numerical values of PPG raw data for each encrypted communication, and the decrypted PPG data are further applied for time‐ and frequency‐domain analyses in clinical applications, such as heart and respiration rate analysis and arterial stiffness and upper extremity vascular disease evaluations. Through experimental results, the security levels are validated using the number of pixel change rate (NPCR), unified averaged changed intensity (UACI), and correlation analysis. The average NPCR, UACI, and correlation coefficient (CC) are 96.57%, 35.43%, and 0.005, respectively, between the plain PPG and the encrypted PPG data against hacker attacks. The larger‐the‐better of NPCR and UACI indexes and the smaller‐the‐better of CC index are obtained to evaluate the efficiency of the proposed cryptography method. The encrypted PPG also guarantees physiological signals of good quality in clinical applications. In addition, the performance of RBFN‐based method is superior in adaptive learning capability than that of the traditional learning method in real‐time applications. The cover image is based on the Original Research Paper Intelligent physiological signal infosecurity: Case study in photoplethysmography (PPG) signal by Chia Hung‐Lin et al., https://doi.org/10.1049/sil2.12089.

Keywords