A decision support system for multi-target disease diagnosis: A bioinformatics approach

Area of Science:

• Bioinformatics
• Computational Biology
• Medical Informatics

Background:

• Malaria and typhoid fever present overlapping symptoms, leading to diagnostic challenges and potential under-diagnosis.
• Co-occurrence of malaria and typhoid fever complicates accurate disease identification.
• High mortality rates are associated with both individual and combined infections.

Purpose of the Study:

• To develop a Bioinformatics Based Decision Support System (BBDSS) for accurate diagnosis of malaria, typhoid fever, and their co-infection.
• To enhance diagnostic accuracy and efficiency by leveraging bioinformatics and expert system approaches.
• To address the diagnostic challenges posed by similar symptoms and co-existence of these diseases.

Main Methods:

• The proposed BBDSS integrates an expert system with a global alignment technique using constant penalty.
• Input diagnosis sequences and benchmark sequences are stored in a knowledge base with IF-THEN rules.
• A matching engine applies global alignment to compare input sequences against benchmark sequences for disease determination.

Main Results:

• Statistical analysis (ANOVA, multiple comparisons) confirmed significant differences in diagnosis variables between disease groups, particularly between malaria and malaria-typhoid.
• T-test statistics showed that the proposed system's diagnosis means differ from the orthodox system.
• The BBDSS demonstrated high diagnostic efficiency, achieving 97% accuracy for malaria and malaria-typhoid diagnoses.

Conclusions:

• The developed BBDSS offers a significant improvement over traditional diagnostic methods for malaria, typhoid, and co-infections.
• The system effectively distinguishes between malaria and malaria-typhoid cases with high accuracy.
• Bioinformatics approaches can successfully enhance clinical decision support for complex infectious diseases.