DiaTTroD: A Logical Agent Diagnostic Test for Tropical Diseases

Abstract

Medical diagnosis is one of the critical areas in medicine. Diagnosing tropical diseases can be confusing if their signs and symptoms are similar. This paper presents a formal logic for constructing a diagnostic test capable of guiding patient examination. A logical agent based on morphological data is used to aid diagnostic procedures and decision-making. To ensure that the appropriate diagnostic procedure is undertaken, the signs and symptoms of the patient are examined first before deciding which exact laboratory examination is needed by the patient. The logical agent perceives the signs, symptoms, medical history, and environment of the patient. Its actuation includes request for laboratory examination. A test kit result can be used, if available, to further confirm the diagnosis. The decision is not based on statistical inference but on logical analysis of the perceived data. Since not all signs and symptoms are present at a certain point in time, using this logical agent will aid the user in diagnosing the patient. A developed test case is presented and result is shown. Test results show 100% accuracy for diseases present in the knowledge base. Also, this paper shows the importance of using morphology in correctly diagnosing a disease. Digital image processing, if completely embedded in this logical agent, will guide the agent in correctly identifying the disease.