Scholarlink

<p>This study introduces a novel hybrid framework for classifying COVID-19 severity using chest X-rays (CXR) and computed tomography (CT) scans by integrating Vision Transformers (ViT) with metaheuristic optimization techniques. The framework employs the Grey Wolf Optimizer (GWO) for hyperparameter tuning and Particle Swarm Optimization (PSO) for feature selection, leveraging the ViT model’s self-attention mechanism to extract global and local image features crucial for severity classification. A multi-phase classification strategy refines predictions by progressively distinguishing normal, mild, moderate, and severe COVID-19 cases. The proposed GWO_ViT_PSO_MLP model achieves outstanding accuracy, with 99.14% for 2-class CXR classification and 98.89% for 2-class CT classification, outperforming traditional CNN-based approaches such as ResNet34 (84.22%) and VGG19 (93.24%). Furthermore, it demonstrates superior performance in multi-class severity classification, especially in differentiating challenging cases like mild and moderate infections. Compared to existing studies, this framework significantly improves accuracy and computational efficiency, highlighting its potential as a scalable and reliable solution for automated COVID-19 severity detection in clinical applications.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Metaheuristic optimizers integrated with vision transformer model for severity detection and classification via multimodal COVID-19 images

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年4月22日

查看全文
  • V. Padmavathi,
  • Kavitha Ganesan

摘要

This study introduces a novel hybrid framework for classifying COVID-19 severity using chest X-rays (CXR) and computed tomography (CT) scans by integrating Vision Transformers (ViT) with metaheuristic optimization techniques. The framework employs the Grey Wolf Optimizer (GWO) for hyperparameter tuning and Particle Swarm Optimization (PSO) for feature selection, leveraging the ViT model’s self-attention mechanism to extract global and local image features crucial for severity classification. A multi-phase classification strategy refines predictions by progressively distinguishing normal, mild, moderate, and severe COVID-19 cases. The proposed GWO_ViT_PSO_MLP model achieves outstanding accuracy, with 99.14% for 2-class CXR classification and 98.89% for 2-class CT classification, outperforming traditional CNN-based approaches such as ResNet34 (84.22%) and VGG19 (93.24%). Furthermore, it demonstrates superior performance in multi-class severity classification, especially in differentiating challenging cases like mild and moderate infections. Compared to existing studies, this framework significantly improves accuracy and computational efficiency, highlighting its potential as a scalable and reliable solution for automated COVID-19 severity detection in clinical applications.