AI-Driven Intelligent Medical System: Deep Learning for Chest X-Ray/CT Disease Recognition | ||||
| Damanhour Journal of Intelligent Systems and Informatics | ||||
| Volume 1, Issue 1 - Serial Number 20240100, January 2025 PDF (981.36 K) | ||||
| Document Type: Original Article | ||||
| DOI: 10.21608/djis.2025.349439.1003 | ||||
 View on SCiNiTO | ||||
| Authors | ||||
Ahmed Saleh1; Ahmed Mohamed Sied2; Rahma Gharib2; Alaa Elsaid2; Ahmed Mabrouk2; Hoda Atef Elbatrawy  2
| ||||
| 1Department of computer science, Faculty of computers and information , Damanhour University, Egypt | ||||
| 2Department of Information Technology, Faculty of Computers and Information, Damanhur University. | ||||
| Abstract | ||||
| In the wake of the global health crisis, healthcare institutions, including hospitals, physicians, medical staff, and patients, faced an urgent need for an advanced medical management system to streamline operations and enhance diagnostic accuracy and efficiency. Consequently, the proposed intelligent medical Chest system (IMCS) leverages artificial intelligence (AI) and deep learning technologies to optimize workflows among healthcare professionals, enabling them to perform their duties more effectively and expedite patient diagnoses with greater precision. The system incorporates both low-dose computed tomography (CT) and chest radiography (CXR) for the screening of lung cancer. While CT offers superior diagnostic precision, it is accompanied by challenges in resource allocation and potential radiation risks. In contrast, CXR serves as a more cost-effective and resource-efficient preliminary screening modality. Leveraging advanced artificial intelligence (AI) and deep learning techniques, the system analyzes both imaging types, streamlining clinical workflows, augmenting diagnostic accuracy, and accelerating patient evaluation and management. By leveraging the sophisticated capabilities of both chest X-rays and CT scans, which each provide unique insights into tissue anomalies, the suggested model dramatically increases diagnostic precision. To efficiently handle and interpret the data from each imaging modality, the model is built on a complex convolutional neural network (CNN) architecture that includes numerous convolutional blocks and fully linked layers. With a classification accuracy of 98.9% for CT scans and 94.6% for X-rays, the system surpasses conventional manual and computerized methods, providing a more thorough and dependable approach for early disease detection and diagnosis. | ||||
| Keywords | ||||
| Chest CT; Image Pre-Processing; Optimization; Deep Learning; Cross-Entropy | ||||
|
Statistics Article View: 318 PDF Download: 146 |
||||
