Mobile-based interpreter of arterial blood gases using knowledge-based expert system

Author： Al-Taee Majid Zayed Ayman Z. Abood Suhail N. Al-Ani Mohammad A. Al-Taee Ahmad M. Hassani Hussein A.

ISSN： 1742-7371

Source： International Journal of Pervasive Computing and Communications, Vol.9, Iss.3, 2013-08, pp. : 270-288

Disclaimer: Any content in publications that violate the sovereignty, the constitution or regulations of the PRC is not accepted or approved by CNPIEC.

Previous Menu Next

Abstract

Purpose - An arterial blood gas (ABG) interpretation remains indispensable tool to assess and monitor critically ill patients in the intensive care unit or other critical care settings. This paper proposes a mobile-based interpreter for ABG tests with the aim of providing accurate diagnosis in face of multiple acid-base and oxygenation disorders. The paper aims to discuss these issues. Design/methodology/approach - A rule-based expert system is designed and implemented using interpretation knowledge gathered from specialist physicians and peer-reviewed medical literature. The gathered knowledge of ABG tests are organized into premise-explanation pairs to deliver reliable evaluation with the appropriate differential in a timely manner. Findings - Performance of the developed interpreter prototype was assessed using a dataset of 74 ABG tests gathered from medical literature and clinical practice. The obtained results demonstrated that the identified acid-base and oxygenation disorders and their differential diagnoses are accurately correlated with those assessed manually by consultant specialist physicians. Research limitations/implications - This application is foreseen to be an everyday tool for clinicians at various levels; however, further studies are needed to evaluate its eventual impact on patients' outcomes. Originality/value - The contribution of this paper is the development of a new ABG interpreter which combines both the acid-base and oxygenation disorders in a single application. Unlike existing ABG interpreters, it is comprehensive and capable of accurately identifying all kinds of acid-base disorders and their combinations. In addition, it utilizes urine electrolytes which are useful tools in the differential diagnosis of normal anion gap metabolic acidosis and metabolic alkalosis. The interpretation algorithm is also designed to be flexible for some clinical settings which lack some input test data.