Author : Jerick Lee 1
Date of Publication :24th June 2021
Abstract: The COVID-19 pandemic placed healthcare systems of every country under immense pressure. Diagnostics and treatment are pushed to its limits as medical frontliners struggle to manage the huge influx of incoming patients with respiratory symptoms. This study aims to assist diagnostics through pre-assessment of X-ray images to detect signals or features that strongly correlates to pneumonia. Specifically, we will train classification neural network on top of various pre-trained Deep Classification Models through existing X-Ray images with and without pneumonia. These models include VGG16, InceptionResNetV2, and MobileNetV2. To test the detection accuracy of each trained model, 25% of the training data will be separated, and will be evaluated after the model has been trained with the remaining images. All images in the dataset are pre-classified, and we will be able to generate accuracy metrics from the evaluation.
Reference :
-
- Aditi, R., Ioannis A., Amar, K.A., Dmitriy, F., Arquimedes, C., Kaushik, K., Tugba, K., (2020). Diag2graph: Representing Deep Learning Diagrams In Research Papers As Knowledge Graphs.IEEEXplore.DOI:10.1109/ICIP40778.2020.919 1234. 2020 IEEE International Conference on Image Processing (ICIP). Abu Dhabi, UAE.
- Ali Bou, I., Atinan, A., Mohammad, A., Khaled, S.,(2019, February 01). Speech Recognition Using Deep Neural Networks: A Systematic Review. IEEE Access. Volume: 7. 19143 – 19165. DOI: 10.1109/ACCESS.2019.2896880. IEEE.
- https://www.kaggle.com/paultimothymooney/chestxray-pneumonia.
- https://keras.io/api/applications.
- Jianyu, W., Zhenling , M., Heng, Z., Qiang, M., (2019, March 27). Advances in Prognostics and System Health Management. IEEE Access. Volume 7. 42373 – 42383. DOI: 10.1109/ACCESS.2019.2907131 IEEE.
- Justin, K., Lipo , W., Jai R.,Tchoyoson L., (2017, December 29). Soft Computing Techniques for Image Analysis in the Medical Industry Current trends, Challenges and Solutions. Volume 6. 9375 – 9389. DOI: 10.1109/ACCESS.2017.2788044. IEEE.
- Ziwei, Z., Peng, C., Wenwu, Z., (2020, March 17). IEEE Transactions on Knowledge and Data Engineering. DOI: 10.1109/TKDE.2020.2981333. IEEE.