ĐÁNH GIÁ KHẢ NĂNG PHÁT HIỆN TỔN THƯƠNG TRÊN X-QUANG NGỰC BẰNG CÔNG CỤ QURE.AI Ở BỆNH NHÂN VIÊM PHỔI CỘNG ĐỒNG

Abstract

Background: The application of Artificial Intelligence (AI) in diagnostic imaging is increasingly widespread and has become a major trend in modern medicine. Pneumonia is a common infectious disease associated with high mortality. The use of AI in analyzing chest X-rays helps in early detection and improves diagnostic accuracy. Objectives: To evaluate the value of the Qure.ai tool in analyzing chest X-ray images in patients with community-acquired pneumonia. Materials and methods: A cross-sectional descriptive study was conducted on 311 patients with community-acquired pneumonia admitted to Can Tho Central General Hospital and Can Tho University of Medicine and Pharmacy Hospital from June 2025 to January 2026. Chest X-ray images were analyzed using the Qure.ai tool and compared with the reference standard. The kappa coefficient, sensitivity, specificity, and diagnostic values were calculated using SPSS 27.0 software. Results: The mean age of the patients was 71.3 ± 13.4 years. The level of agreement between Qure.ai and radiologists in detecting pneumonia-related lesions ranged from moderate to high. Pleural effusion lesions showed the highest level of agreement (κ = 0.708). In the subgroup of patients who underwent chest computed tomography as the gold standard, Qure.ai detected consolidation/opacity lesions with a sensitivity of 91.7%. Among the 60 cases with confirmed lesions on chest CT, the Qure.ai tool exhibited higher sensitivity compared to radiologists (98.3% versus 91.7%), though the difference was not statistically significant (p = 0.125). Conclusion: Qure.ai demonstrates significant potential as a supportive tool for detecting pneumonic lesions on chest radiographs.