Detection of Viral Pneumonia in Radiographic Images Using Deep Learning
Issued Date
2024-01-01
Resource Type
Scopus ID
2-s2.0-85209665804
Journal Title
2024 5th International Conference on Artificial Intelligence and Data Sciences, AiDAS 2024 - Proceedings
Start Page
239
End Page
244
Rights Holder(s)
SCOPUS
Bibliographic Citation
2024 5th International Conference on Artificial Intelligence and Data Sciences, AiDAS 2024 - Proceedings (2024) , 239-244
Suggested Citation
Laesanklang W., Lohajareekul P. Detection of Viral Pneumonia in Radiographic Images Using Deep Learning. 2024 5th International Conference on Artificial Intelligence and Data Sciences, AiDAS 2024 - Proceedings (2024) , 239-244. 244. doi:10.1109/AiDAS63860.2024.10730357 Retrieved from: https://repository.li.mahidol.ac.th/handle/123456789/102183
Title
Detection of Viral Pneumonia in Radiographic Images Using Deep Learning
Author(s)
Corresponding Author(s)
Other Contributor(s)
Abstract
Viral pneumonia, often caused by viruses such as COVID-19, is a significant infection of the lungs. Detection of this condition can be achieved through radiographic imaging techniques, including chest CT and X-ray images. Previous research has demonstrated the effectiveness of neural networks in detecting symptoms of viral pneumonia. However, the model's performance declines when the input images suffer from issues such as varying positions, rotations, and inconsistent scales. In this research, we employ deep learning models to diagnose lung diseases from chest X-ray images. To enhance symptom classification by mitigating the effects of image positioning, we incorporate a Spatial Transformer Network (STN). This STN technique is integrated with Convolutional Neural Networks (CNN) and ResNet architectures. Our experimental results demonstrate that the inclusion of the Spatial Transformer Network significantly improves the classification performance of the deep learning models.