Related Experiment Video
Updated: Aug 5, 2025
08:05
Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia
Published on: December 19, 2020
14.2K
MTMC-AUR2CNet: Multi-textural multi-class attention recurrent residual convolutional neural network for COVID-19
Anandbabu Gopatoti1,2, P Vijayalakshmi1
1Department of Electronics and Communication Engineering, Hindusthan College of Engineering and Technology, Coimbatore, Tamil Nadu, India.
Biomedical Signal Processing and Control
|March 27, 2023
Summary
Deep learning models, MTMC-AUR2CNet and MTMC-UR2CNet, accurately segment lung lobes in chest X-rays for COVID-19 diagnosis. MTMC-AUR2CNet achieved 99.47% accuracy in segmentation and 97.60% in classification.
Area of Science:
- Artificial Intelligence
- Medical Imaging
- Deep Learning
Background:
- The COVID-19 pandemic has caused millions of deaths globally, with limited diagnostic tools.
- Deep learning (DL) models offer a promising approach to assist healthcare professionals in diagnosing COVID-19 using chest X-ray (CXR) images.
Purpose of the Study:
- To develop and evaluate novel deep learning models for multi-class lung lobe segmentation and classification of CXR images.
- To improve the accuracy and efficiency of COVID-19 diagnosis through advanced image analysis techniques.
Main Methods:
- Proposed two UNet-based Recurrent Residual Convolutional Neural Networks: MTMC-UR2CNet and MTMC-AUR2CNet with an attention mechanism.
- Implemented multi-textural feature extraction from regions of interest (ROIs) derived from lung lobe segmentation.
- Utilized a Whale Optimization Algorithm (WOA)-based DeepCNN classifier for multi-class classification (normal, COVID-19, viral pneumonia, lung opacity).
Main Results:
- MTMC-AUR2CNet demonstrated superior performance in multi-class lung lobe segmentation with 99.47% accuracy, outperforming MTMC-UR2CNet (98.39%).
- MTMC-AUR2CNet also enhanced the multi-textural multi-class classification accuracy to 97.60%, compared to MTMC-UR2CNet.
Conclusions:
- The proposed MTMC-AUR2CNet model shows significant potential for accurate COVID-19 diagnosis using CXR images.
- Attention mechanisms in DL models can substantially improve performance in medical image analysis tasks.
