• DocumentCode
    1202262
  • Title

    Assessing cardiovascular dynamics during ventricular assistance. Use of fuzzy clustering techniques

  • Author

    Yoshizawa, Makoto ; Takeda, Hiroshi ; Yambe, Tomoyuki ; Nitta, Shin-ichi

  • Author_Institution
    Graduate Sch. of Inf. Sci., Tohoku Univ., Sendai, Japan
  • Volume
    13
  • Issue
    5
  • fYear
    1994
  • Firstpage
    687
  • Lastpage
    692
  • Abstract
    The authors´ group has recently developed an expert system realized by means of fuzzy logic algorithms for monitoring and malfunction diagnosis of the left ventricular assist device (LVAD), one of the many types of artificial hearts developed today. The system is named TOTOMES (Tohoku University and Toyohashi University of Technology, monitoring and estimation system). TOTOMES is suitable for clinical application because it operates in an online and realtime fashion on a widely used personal computer in Japan. This article introduces the structure and function of the TOTOMES and explains its techniques, such as multi-interrupt tasks and dynamic system identification; as well as fuzzy reasoning used for realizing state estimation, detection, and diagnosis of malfunctions; and monitoring for cardiovascular dynamics under ventricular assistance and the LVAD drive system.<>
  • Keywords
    fuzzy logic; haemodynamics; medical expert systems; TOTOMES; Tohoku University; Toyohashi University of Technology; cardiac ventricular assistance; cardiovascular dynamics assessment; dynamic system identification; fuzzy clustering techniques; fuzzy logic algorithms; left ventricular assist device malfunction diagnosis; left ventricular assist device monitoring; multiinterrupt tasks; state estimation; Application software; Artificial heart; Cardiology; Computerized monitoring; Diagnostic expert systems; Fuzzy logic; Fuzzy reasoning; Microcomputers; State estimation; System identification;
  • fLanguage
    English
  • Journal_Title
    Engineering in Medicine and Biology Magazine, IEEE
  • Publisher
    ieee
  • ISSN
    0739-5175
  • Type

    jour

  • DOI
    10.1109/51.334630
  • Filename
    334630