Related Experiment Video
Updated: Oct 9, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
3.0K
Effect of image transformation on EfficientNet model for COVID-19 CT image classification
A Shamila Ebenezer1, S Deepa Kanmani2, Mahima Sivakumar1
1Department of Computer Science and Engineering, Karunya Institute of Technology and Sciences, Coimbatore, Tamil Nadu 641114, India.
Materials Today. Proceedings
|December 20, 2021
Summary
This study enhanced COVID-19 detection using Chest CT scans with image enhancement techniques. Contrast Limited Adaptive Histogram Equalization (CLAHE) with EfficientNet achieved 94.56% accuracy, improving diagnostic performance.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Computer Vision
Background:
- The Novel Coronavirus (COVID-19) pandemic poses a significant global health threat.
- Chest CT scans are crucial for diagnosing COVID-19.
- Artificial Intelligence (AI) models show promise in analyzing medical images for disease detection.
Purpose of the Study:
- To evaluate the impact of various image enhancement algorithms on COVID-19 classification accuracy using Chest CT images.
- To assess the performance of the EfficientNet Convolutional Neural Network (CNN) model when combined with different preprocessing techniques.
Main Methods:
- Utilized the SARS-COV-2 dataset comprising Chest CT images.
- Applied image enhancement techniques: Laplace transform, Wavelet transforms, Adaptive gamma correction, and Contrast Limited Adaptive Histogram Equalization (CLAHE).
- Implemented and evaluated the EfficientNet model with and without these enhancement algorithms.
Main Results:
- The EfficientNet model integrated with CLAHE achieved the highest performance metrics.
- CLAHE-Enhanced EfficientNet yielded an accuracy of 94.56%, precision of 95%, recall of 91%, and F1-score of 93%.
- Image enhancement significantly improved the classification performance of the EfficientNet model.
Conclusions:
- Contrast Limited Adaptive Histogram Equalization (CLAHE) is an effective preprocessing step for improving COVID-19 classification from Chest CT images using EfficientNet.
- Integrating image enhancement techniques with deep learning models like EfficientNet can enhance diagnostic accuracy for infectious diseases.
