Healthcare Analytics (Nov 2022)
A fuzzy model for predicting burn patients’ intravenous fluid resuscitation rate
Abstract
One of the critical areas in the medical field is the burn treatment management that helps to rescue patient’s life by dramatically minimizing the after-burn progression and severity. Initial acute burn treatment involves intravenous fluid resuscitation, which must be administered meticulously to restore fluid loss and maintain essential organ function. A number of factors must be addressed quickly to make a decision for determining the required rate of intravenous fluid. Traditional methods with manual calculation reflects slow response and inaccurate estimation in most of the cases, thus requires random adjustments. Some automated and artificially intelligent systems can be found which still use the traditional formulas showing high fluctuation in the calculation of required rate. Therefore, this research proposes a fuzzy logic based model to predict the adequate intravenous fluid resuscitation rate for a burn patient. Fuzzy systems are treated as intelligent system close to human reasoning capabilities and for this application it ensures smooth transition of required rate without showing any high fluctuations for any adjustment. To attain the objective, a fuzzy logic system has been developed aggregating the clinical burn protocol as the knowledge base of the fuzzy inference engine where the %TBSA and HUO measurements are considered as inputs and intravenous fluid rate is the output. Multiple forms of fuzzy membership functions and different types of defuzzification methods have been applied to the proposed system to compile a comparative study. To explore the best approach, ten different test cases have been simulated employing both the manual calculation and the proposed fuzzy model. Results have presented that the proposed fuzzy based system shows to predict the optimum amount of intravenous fluid rate necessary for burn patient’s resuscitation.
