Fault Prioritization System for a Vital Signs Monitor Based on Fuzzy Logic

Abstract

This work presents a fault prioritization system based in two MISO (Multi Input Single Output) fuzzy logic systems, applied on a vital signs monitor. The MISO systems correspond to possible failures, classified as physiological and technical failure. For physiological failures, the fuzzy inputs are heart rate, respiratory rate, temperature and the partial oxygen saturation variables, the membership functions for each variable were designed using the normal range for a healthy adult. Technical failures system considered faults in the monitor’s arrange of sensors as inputs, while it’s membership functions vary accordingly to the operational status of the sensors. Both fuzzy systems have as output four priority levels, whose labels are: normal, mild, moderate and critical. Once the design of the MISO systems was carried out, three trials were performed on the fault prioritization system: a Monte Carlo test for both subsystems, simulated patient test for physiological failures and real patient with sensors connectivity alterations for technical failures. Due to the results achieved, the fault prioritization system can be applied in the automation of alarms or can be a reference for the development of future Fuzzy systems applied to different types of vital signs monitors.