What is it about?
• The first attempt at CXR image classification with multi-textural features and multi-class lung lobe segmentation was introduced. • The MTMC-UR2CNet and MTMC-AUR2CNet are developed for multi-class lung lobes segmentation of CXR images. • Lung lobes segmentation output is mapped with input CXRs to obtain the ROI. • ROI is used to extract multi-textural features to improve multi-class classification. • A whale optimization algorithm (WOA)-based DeepCNN classifier is developed to classify the CXR images into normal (healthy), COVID-19, viral pneumonia, and lung opacity using extracted multi-textural features.
Featured Image
Why is it important?
The outcome of the proposed multi-textural multi-class DL models for COVID-19 diagnosis from CXR images is promising and encouraging, suggesting that deep learning will soon play a more significant role in the fight against the COVID-19 pandemic. We hope that the proposed deep models in this work and their findings will be a starting point for developing a technique that uses multi-textural features for chest X-ray chest image-based diagnosis of chest diseases.
Perspectives
Current research aims to improve the performance of the suggested models by increasing the number of images in the datasets, increasing the training epochs, increasing the number of classes, and making the models more suitable in all CXR classes.
Anandbabu Gopatoti
Anna University Chennai
Read the Original
This page is a summary of: MTMC-AUR2CNet: Multi-textural multi-class attention recurrent residual convolutional neural network for COVID-19 classification using chest X-ray images, Biomedical Signal Processing and Control, August 2023, Elsevier,
DOI: 10.1016/j.bspc.2023.104857.
You can read the full text:
Contributors
The following have contributed to this page
