Modelling Differential Diagnosis of Febrile Diseases with Fuzzy Cognitive Map

Area of Science:

• Medical Informatics
• Computational Medicine
• Public Health

Background:

• Limited access to healthcare in low-to-middle-income countries contributes to high morbidity and mortality.
• Febrile diseases with overlapping symptoms pose diagnostic challenges for inexperienced clinicians.

Purpose of the Study:

• To develop a Fuzzy Cognitive Map (FCM) model as a decision-support tool for diagnosing febrile diseases.
• To improve diagnostic accuracy and healthcare accessibility in resource-limited tropical regions.

Main Methods:

• Gathered 2465 datasets from febrile disease-prone regions in Nigeria.
• Utilized input from 60 medical doctors, with 10 experts weighting and fuzzifying symptoms.
• Generated an FCM model based on weighted and fuzzified symptom data.

Main Results:

• The FCM model demonstrated an average accuracy of 87% in diagnosing 11 types of febrile diseases.
• The model's predictions for comorbidities and disease severity strongly correlated with physician diagnoses.
• The study highlights the potential of AI in enhancing diagnostic capabilities in underserved areas.

Conclusions:

• The developed Fuzzy Cognitive Map (FCM) model is an effective decision-support tool for febrile disease diagnosis.
• This technology can significantly improve diagnostic accuracy and patient outcomes in regions with limited medical resources.
• The FCM approach offers a scalable solution to address healthcare accessibility challenges reported by the WHO.