Pilot Expert System for the diagnosis of the health status of patients with covid-19 based on vital signs

Abstract

This article presents a fuzzy pilot expert system as a complementary alternative measure with the objetive to quickly diagnose the health status of a patient who may have or develop complications associated with Covid-19. For the case of this research, a mixed descriptive methodology was used. In the first instance, an Expert System (ES) was developed to collect information on Covid-19, later, through a semantic network, the knowledge about this disease was visualized so that, from the declaration of knowledge to natural language, define the variables and perform a parameterization process for the creation of the Fuzzy System (FS). For the design of the questionnaire items, a medical specialist was consulted and the instrument was applied through a Google form to a random population of 72 patients who had presented the symptoms of Covid-19. The results obtained by applying the ES showed an efficiency of 86% in the diagnosis made to the population sample, justifying the supposed hypothesis of a diagnosis greater than 80%. It is concluded that since Covid-19 is a pandemic disease with variants and manifestation of multiple symptoms in patients who have acquired it, the generation of new diagnostic methodologies such as the one presented here, allow a rapid diagnosis in the face of the high demand for medical care. This diagnostic modality through an ES, to the extent that feedback continues and is applied to a greater number of patients, will help to detect more efficiently and in time whether or not the patient has Covid-19.