<![CDATA[Lung disease remains a major health problem in Indonesia, accounting for 25.8% of respiratory-related deaths according to the Ministry of Health. Data from Muhammad Ali Kasim Gayo Lues Regional Hospital, Aceh, shows approximately 3,600 cases recorded since 2022. This study designs an artificial intelligence-based diagnostic system combining 224x224 pixel chest X-ray image analysis with clinical parameter evaluation using a rule-based system. The Rule-Based System implements standardized weighting where each clinical manifestation contributes proportionally to the total diagnostic score through normalization to a 100-point scale.The dataset consists of 983 images (786 training, 197 validation) collected during the 2022-2025 period, covering tuberculosis (300 cases), pneumonia (300 cases), and pneumothorax (383 cases) with 25 disease symptoms. Evaluation results show the system achieves 94.97% accuracy with 93.2% sensitivity and 95.4% specificity. The F1-scores for each disease were 0.9375 (Tuberculosis), 0.9125 (Pneumonia), and 0.987 (Pneumothorax). Therefore, this system can assist the diagnostic process and support clinical decision-making through both radiographic image analysis and patient symptom evaluation.]]>
Copyrights © 2025
