Scholarlink

<p>Accurate and early detection of pulmonary illnesses, including pneumonia and COVID-19, is crucial for effective treatment and improved patient outcomes. This study presents a novel DconvNET-based deep learning framework integrated with hybrid clustering (K-Means + Fuzzy C-Means) to enhance lung disease classification using chest X-ray images. Unlike conventional methods, this approach significantly improves segmentation accuracy while reducing computational complexity. The proposed methodology follows a multi-step process, beginning with image acquisition, preprocessing, hybrid segmentation, feature extraction, and classification using DconvNET. Hybrid clustering improves region-of-interest (ROI) extraction, allowing for more precise lung disease segmentation compared to state-of-the-art methods such as Deeplab, SegNet, and FCN. Meanwhile, DconvNET ensures high-dimensional feature learning, optimizing classification accuracy. The dataset used for model training consists of 15,000 chest X-ray images from publicly available repositories. Experimental results demonstrate the superiority of our approach, achieving 0.99723 accuracy, 0.98970 precision, and 0.99875 specificity, outperforming traditional CNN-K-Means, CNN-C-Means, and SVM-based models. The proposed framework not only enhances segmentation and classification accuracy but also significantly reduces processing time by 28%, making it feasible for real-time clinical implementation. Our findings highlight the potential of AI-driven hybrid techniques in revolutionizing lung disease diagnosis, providing a robust and efficient tool for automated medical imaging analysis.</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

DconvNET Based Detection and Classification of Lung Diseases for Pneumonia

  • 期刊论文

  • 发表日期：

    2025年10月27日

查看全文
  • P. V. Naga Lakshmi,
  • K. Vedavathi

摘要

Accurate and early detection of pulmonary illnesses, including pneumonia and COVID-19, is crucial for effective treatment and improved patient outcomes. This study presents a novel DconvNET-based deep learning framework integrated with hybrid clustering (K-Means + Fuzzy C-Means) to enhance lung disease classification using chest X-ray images. Unlike conventional methods, this approach significantly improves segmentation accuracy while reducing computational complexity. The proposed methodology follows a multi-step process, beginning with image acquisition, preprocessing, hybrid segmentation, feature extraction, and classification using DconvNET. Hybrid clustering improves region-of-interest (ROI) extraction, allowing for more precise lung disease segmentation compared to state-of-the-art methods such as Deeplab, SegNet, and FCN. Meanwhile, DconvNET ensures high-dimensional feature learning, optimizing classification accuracy. The dataset used for model training consists of 15,000 chest X-ray images from publicly available repositories. Experimental results demonstrate the superiority of our approach, achieving 0.99723 accuracy, 0.98970 precision, and 0.99875 specificity, outperforming traditional CNN-K-Means, CNN-C-Means, and SVM-based models. The proposed framework not only enhances segmentation and classification accuracy but also significantly reduces processing time by 28%, making it feasible for real-time clinical implementation. Our findings highlight the potential of AI-driven hybrid techniques in revolutionizing lung disease diagnosis, providing a robust and efficient tool for automated medical imaging analysis.