COVID-19 Detection from Chest X-rays Using Trained Output Based Transfer Learning Approach

Abstract

The recent Coronavirus disease (COVID-19), which started in 2019, has spread across the globe and become a global pandemic. The efficient and effective COVID-19 detection using chest X-rays helps in early detection and curtailing the spread of the disease. In this paper, we propose a novel Trained Output-based Transfer Learning (TOTL) approach for COVID-19 detection from chest X-rays. We start by preprocessing the Chest X-rays of the patients with techniques like denoising, contrasting, segmentation. These processed images are then fed to several pre-trained transfer learning models like InceptionV3, InceptionResNetV2, Xception, MobileNet, ResNet50, ResNet50V2, VGG16, and VGG19. We fine-tune these models on the processed chest X-rays. Then we further train the outputs of these models using a deep neural network architecture to achieve enhanced performance and aggregate the capabilities of each of them. The proposed model has been tested on four recent COVID-19 chest X-rays datasets by computing several popular evaluation metrics. The performance of our model has also been compared with various deep transfer learning models and several contemporary COVID-19 detection methods. The obtained results demonstrate the efficiency and efficacy of our proposed model.

Dynamic learning for imbalanced data in learning chest X-ray and CT images

“…The second study involved the use of several pre-trained CNN models, including DenseNet121, for classifying COVID-19 cases based on mixed datasets of chest X-ray (CXR) and CT images. The outcomes showed that DenseNet121 had the advisable achievement, with an accuracy of 0.99 [67] , [68] . These studies demonstrate the potential of deep learning techniques and radiomics for predicting patient outcomes in COVID-19 cases and may be useful for identifying patients who are at higher risk and in need of more intensive treatment.…”

Generalizable disease detection using model ensemble on chest X-ray images

“…By leveraging pre-trained model architectures, transfer learning enables faster training processes with fewer input data, while improving overall model efficiency and generalization. This approach has significantly contributed to advancements in medical image artificial intelligence applications, as researchers and practitioners have successfully applied pre-trained models to enhance diagnostic capabilities across various medical imaging domains 9 , 10 .…”

The Impact of Simultaneous Adversarial Attacks on Robustness of Medical Image Analysis

“…In addition, it contains a fully connected layer for classification after a global average pooling layer. The depthwise separable convolutional layers comprise a depthwise convolution operation that uses one filter for each input channel and a pointwise convolution operation that combines the results of the depthwise convolution to create the final output[30].• Alexnet: It has eight layers, five of which are convolutional and three are completely connected. Each convolutional layer's output is subjected to the network's Rectified Linear Unit (ReLU) activation function.…”