Next Generation Chest X-Ray Classification with Effitune-B1: Optimized Transfer Learning for Respiratory Health Screening
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
The COVID-19 outbreak has highlighted the worldwide demand for intelligent, data-driven systems that enhance timely analysis and support effective healthcare decision-making. Deep learning has become a key enabler in medical imaging, particularly for chest X-ray analysis, due to its accessibility and diagnostic relevance. However, manual interpretation remains challenging and resource-intensive. This study presents Effitune-B1, an improved form of the EfficientNet model, refined using a unified scaling strategy that proportionally adjusts the model's layers, feature dimensions, and input size. This coordinated scaling approach enables the network to achieve high accuracy while keeping computational demands low. A transfer learning framework built on the Effitune-B1 model was designed to categorize chest X-ray images into four classes: COVID-19, lung opacity, viral pneumonia, and normal. Using the COVID-19 Radiography Dataset for training and validation, the system achieved an accuracy of 95.5% and a kappa value of 95%, exceeding the performance of several existing approaches. These results highlight the capability of Effitune-B1 to support automated, efficient, and scalable analysis of medical images. These findings contribute to the ongoing evolution of AI-assisted healthcare, promoting faster, reliable, and cost-effective detection of respiratory diseases.
Fields of Study & Subjects
COVID-19 diagnosis using AI
Advanced Neural Network Applications
Artificial Intelligence in Healthcare and Education
Publication details
- Journal / source
- International Conference on Intelligent Systems for Communication, IoT and Security (ICISCoIS)
- Pages
- 1–7
- Publication date
- 8 January 2026
- Language
- EN