作者： D. A. Linkens , K. M. Goode , P. R. Bourne , J. G. Cundill
DOI:
关键词: Medicine 、 Mechanical ventilation 、 Fuzzy rule 、 Blood pressure 、 Arterial pH 、 Surgery 、 Patient data 、 Tidal volume 、 Patient model 、 Emergency medicine 、 Advisory system
摘要: An important aspect of patient care in ICU is the management arterial pH, O 2 partial pressure (PaO ) and CO (PaCO via mechanical ventilation. anaesthetist makes adjustments to ventilator settings, based upon observations current state. The ability mimic this decision making process using fuzzy logic a rule-based advisory system has been investigated. advisor suggests new settings for inspired fraction (FiO ), positive end-expiratory (PEEP), respiratory rate (RR) tidal volume (Vt) PaO , PaCO systolic blood (BP SYS ). A model mechanically ventilated patients was used validate refine rule-base, absence real data. This enabled representation various traumas pathologies, which two examples are presented; an over anaesthetised patient, following gastric resection surgery with lobar pneumonia. Analysis performance when connected simulated closed-loop control, highlighted rule errors rapid modification. modified showed good matching compared chosen by anaesthetist. graded computed advice order establish its acceptability safety. He felt that 75.0 % advised changes were definitely acceptable, 21.9 different but acceptable 3.1 should possibly be rejected. No completely rejected none deemed unsafe. Of decisions differed from those proposed anaesthetist, most gave better resultant gases levels.