Scholarlink

This research presents a novel ensemble deep learning (DL) methodology for identifying various pulmonary diseases (such as: pneumonia, tuberculosis, and Covid-19) from chest X-ray images. This technique ensembled MobileNetV2, ResNet50, InceptionV3, and VGG16 based on their computational parameters and effectiveness in capturing cross channel features for improving performance, robustness, and avoiding overfitting. The technique was experimented on the datasets available in Kaggle (615 Pneumonia, 595 Tuberculosis (TB), 598 COVID-19, and 632 normal). Individually ResNet50, MobileNetV2, VGG16, and InceptionV3 achieved an accuracy of 36.07%, 94.95%, 81.54%, 86.16%, respectively. These models are ensembled by stacking & weighted voting and reported the highest accuracy of 95%.

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

An Ensemble DL Approach for Robust Identification of Pulmonary Diseases in Chest X Ray Images

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Malaya Kumar Nath,
  • Jayashree Natrajan,
  • K. Kirthika

摘要

This research presents a novel ensemble deep learning (DL) methodology for identifying various pulmonary diseases (such as: pneumonia, tuberculosis, and Covid-19) from chest X-ray images. This technique ensembled MobileNetV2, ResNet50, InceptionV3, and VGG16 based on their computational parameters and effectiveness in capturing cross channel features for improving performance, robustness, and avoiding overfitting. The technique was experimented on the datasets available in Kaggle (615 Pneumonia, 595 Tuberculosis (TB), 598 COVID-19, and 632 normal). Individually ResNet50, MobileNetV2, VGG16, and InceptionV3 achieved an accuracy of 36.07%, 94.95%, 81.54%, 86.16%, respectively. These models are ensembled by stacking & weighted voting and reported the highest accuracy of 95%.