A Novel Approach for Prediction of Lung Disease Using Chest X-ray Images Based on DenseNet and MobileNet

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Computer-Aided Diagnosis

Background:

• COVID-19 pandemic caused global health and economic disruptions.
• Chest X-rays are crucial for diagnosing lung manifestations of COVID-19.
• Accurate and timely diagnosis is essential for patient management.

Purpose of the Study:

• To propose a deep learning-based classification method for identifying COVID-19 from chest X-ray images.
• To evaluate the effectiveness of MobileNet and DenseNet models in COVID-19 detection.
• To assess the performance of the proposed method using metrics like accuracy, AUC, precision, recall, and F1-Score.

Main Methods:

• Utilized deep learning models, specifically MobileNet and DenseNet, for image classification.
• Employed a case modeling approach with the MobileNet model.
• Trained and validated the models on a dataset of chest X-ray images.

Main Results:

• Achieved 96% accuracy and 94% Area Under Curve (AUC) using the MobileNet model with case modeling.
• Demonstrated the potential for accurate identification of lung impurities indicative of COVID-19.
• Compared various performance metrics including precision, recall, and F1-Score.

Conclusions:

• The proposed deep learning method shows high accuracy in detecting COVID-19 from chest X-rays.
• MobileNet and DenseNet are effective deep learning models for this diagnostic task.
• This approach can assist healthcare professionals in the accurate diagnosis of COVID-19.