A zero-watermarking algorithm for privacy protection in biomedical signals

Zulfiqar Ali, Muhammad Imran, Mansour Alsulaiman, Tanveer Zia, Muhammad Shoaib

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

Confidentiality of health information is indispensable to protect privacy of an individual. However, recent advances in electronic healthcare systems allow transmission of sensitive information through the Internet, which is prone to various vulnerabilities, attacks and may leads to unauthorized disclosure. Such situations may not only create adverse effects for individuals but may also cause severe consequences such as hefty regulatory fines, bad publicity, legal fees, and forensics. To avoid such predicaments, a privacy protected healthcare system is proposed in this study that protects the identity of an individual as well as detects vocal fold disorders. The privacy of the developed healthcare system is based on the proposed zero-watermarking algorithm, which embeds a watermark in a secret key instead of the signals to avoid the distortion in an audio sample. The identity is protected by the generation of its secret shares through visual cryptography. The generated shares are embedded by finding the patterns into the audio with the application of one-dimensional local binary pattern. The proposed zero-watermarking algorithm is evaluated by using audio samples taken from the Massachusetts Eye and Ear Infirmary voice disorder database. Experimental results demonstrate that the proposed algorithm achieves imperceptibility and is reliable in its extraction of identity. In addition, the proposed algorithm does not affect the results of disorder detection and it is robust against noise attacks of various signal-to-noise ratios.

LanguageEnglish
Pages290-303
Number of pages14
JournalFuture Generation Computer Systems
Volume82
Early online date19 Dec 2017
DOIs
Publication statusPublished - 31 May 2018

Fingerprint

Watermarking
Cryptography
Signal to noise ratio
Health
Internet

Keywords

  • E-healthcare
  • Local binary pattern
  • MFCC
  • Privacy protection
  • SVM
  • Visual cryptography
  • Zero-watermarking

Cite this

Ali, Zulfiqar ; Imran, Muhammad ; Alsulaiman, Mansour ; Zia, Tanveer ; Shoaib, Muhammad. / A zero-watermarking algorithm for privacy protection in biomedical signals. In: Future Generation Computer Systems. 2018 ; Vol. 82. pp. 290-303.
@article{a6bc9046d1394c96a2dbc32f7ccffa07,
title = "A zero-watermarking algorithm for privacy protection in biomedical signals",
abstract = "Confidentiality of health information is indispensable to protect privacy of an individual. However, recent advances in electronic healthcare systems allow transmission of sensitive information through the Internet, which is prone to various vulnerabilities, attacks and may leads to unauthorized disclosure. Such situations may not only create adverse effects for individuals but may also cause severe consequences such as hefty regulatory fines, bad publicity, legal fees, and forensics. To avoid such predicaments, a privacy protected healthcare system is proposed in this study that protects the identity of an individual as well as detects vocal fold disorders. The privacy of the developed healthcare system is based on the proposed zero-watermarking algorithm, which embeds a watermark in a secret key instead of the signals to avoid the distortion in an audio sample. The identity is protected by the generation of its secret shares through visual cryptography. The generated shares are embedded by finding the patterns into the audio with the application of one-dimensional local binary pattern. The proposed zero-watermarking algorithm is evaluated by using audio samples taken from the Massachusetts Eye and Ear Infirmary voice disorder database. Experimental results demonstrate that the proposed algorithm achieves imperceptibility and is reliable in its extraction of identity. In addition, the proposed algorithm does not affect the results of disorder detection and it is robust against noise attacks of various signal-to-noise ratios.",
keywords = "E-healthcare, Local binary pattern, MFCC, Privacy protection, SVM, Visual cryptography, Zero-watermarking",
author = "Zulfiqar Ali and Muhammad Imran and Mansour Alsulaiman and Tanveer Zia and Muhammad Shoaib",
note = "Author started at Ulster Feb-19",
year = "2018",
month = "5",
day = "31",
doi = "10.1016/j.future.2017.12.007",
language = "English",
volume = "82",
pages = "290--303",
journal = "Future Generation Computer Systems",
issn = "0167-739X",
publisher = "Elsevier",

}

A zero-watermarking algorithm for privacy protection in biomedical signals. / Ali, Zulfiqar; Imran, Muhammad; Alsulaiman, Mansour; Zia, Tanveer; Shoaib, Muhammad.

In: Future Generation Computer Systems, Vol. 82, 31.05.2018, p. 290-303.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A zero-watermarking algorithm for privacy protection in biomedical signals

AU - Ali, Zulfiqar

AU - Imran, Muhammad

AU - Alsulaiman, Mansour

AU - Zia, Tanveer

AU - Shoaib, Muhammad

N1 - Author started at Ulster Feb-19

PY - 2018/5/31

Y1 - 2018/5/31

N2 - Confidentiality of health information is indispensable to protect privacy of an individual. However, recent advances in electronic healthcare systems allow transmission of sensitive information through the Internet, which is prone to various vulnerabilities, attacks and may leads to unauthorized disclosure. Such situations may not only create adverse effects for individuals but may also cause severe consequences such as hefty regulatory fines, bad publicity, legal fees, and forensics. To avoid such predicaments, a privacy protected healthcare system is proposed in this study that protects the identity of an individual as well as detects vocal fold disorders. The privacy of the developed healthcare system is based on the proposed zero-watermarking algorithm, which embeds a watermark in a secret key instead of the signals to avoid the distortion in an audio sample. The identity is protected by the generation of its secret shares through visual cryptography. The generated shares are embedded by finding the patterns into the audio with the application of one-dimensional local binary pattern. The proposed zero-watermarking algorithm is evaluated by using audio samples taken from the Massachusetts Eye and Ear Infirmary voice disorder database. Experimental results demonstrate that the proposed algorithm achieves imperceptibility and is reliable in its extraction of identity. In addition, the proposed algorithm does not affect the results of disorder detection and it is robust against noise attacks of various signal-to-noise ratios.

AB - Confidentiality of health information is indispensable to protect privacy of an individual. However, recent advances in electronic healthcare systems allow transmission of sensitive information through the Internet, which is prone to various vulnerabilities, attacks and may leads to unauthorized disclosure. Such situations may not only create adverse effects for individuals but may also cause severe consequences such as hefty regulatory fines, bad publicity, legal fees, and forensics. To avoid such predicaments, a privacy protected healthcare system is proposed in this study that protects the identity of an individual as well as detects vocal fold disorders. The privacy of the developed healthcare system is based on the proposed zero-watermarking algorithm, which embeds a watermark in a secret key instead of the signals to avoid the distortion in an audio sample. The identity is protected by the generation of its secret shares through visual cryptography. The generated shares are embedded by finding the patterns into the audio with the application of one-dimensional local binary pattern. The proposed zero-watermarking algorithm is evaluated by using audio samples taken from the Massachusetts Eye and Ear Infirmary voice disorder database. Experimental results demonstrate that the proposed algorithm achieves imperceptibility and is reliable in its extraction of identity. In addition, the proposed algorithm does not affect the results of disorder detection and it is robust against noise attacks of various signal-to-noise ratios.

KW - E-healthcare

KW - Local binary pattern

KW - MFCC

KW - Privacy protection

KW - SVM

KW - Visual cryptography

KW - Zero-watermarking

UR - http://www.scopus.com/inward/record.url?scp=85040130768&partnerID=8YFLogxK

U2 - 10.1016/j.future.2017.12.007

DO - 10.1016/j.future.2017.12.007

M3 - Article

VL - 82

SP - 290

EP - 303

JO - Future Generation Computer Systems

T2 - Future Generation Computer Systems

JF - Future Generation Computer Systems

SN - 0167-739X

ER -