JISE


  [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]


Journal of Information Science and Engineering, Vol. 39 No. 6, pp. 1305-1319


A Schizophrenia Screening Method Based on Fine-grained Voice Features and Network Isolation Privacy Protection


XIANCHUAN LING1, XIN ZHANG1,+, HUIRU WANG1, GUOZHU ZHAO2,
HUANHUAN ZHAO2, SHENGHUI ZHAO2 AND ZENGHUI DING3
1Chuzhou Second People’s Hospital, Anhui, 239000 P.R. China
E-mail: zhangxin0550@126.com

2Chuzhou University, Anhui, 239000 P.R. China
E-mail: gzzhao2015@126.com

3Institute of Intelligent Machines, Hefei Institutes of Physical Science
Chinese Academy of Sciences, Anhui, 230031 P.R. China
E-mail: dingzenghui@iim.ac.cn


Schizophrenia screening and risk assessment is critical to social harmony and stability as well as personal health and growth. Based on audio data from 179 schizophrenia convalescent patients, this paper proposes a Support Vector Machine (SVM) based schizophrenia screening and risk assessment framework utilizing network physical isolation within a private cloud for user privacy protection. We first design a network security strategy based on wireless sensor network using network physical isolation technology within a private cloud to ensure the absolute security of users’ personal privacy. We further exploit the oneclass classification technique to formulate the schizophrenia screening decision process as an SVM model. Extensive experiments are conducted to illustrate the schizophrenia screening performance of the proposed SVM based screening framework in terms of accuracy. We also investigate the related authentication efficiency issues in terms of the usability to audio contents and the scalability to the number of audio features.


Keywords: schizophrenia screening, wireless sensor networks, SVM, network physical isolation, privacy protection

  Retrieve PDF document (JISE_202306_06.pdf)