Scholarlink

<p>Pneumonia is a virulent disease that has caused deaths of millions of people globally. Each year, it claims more children’s lives than malaria, AIDS, and measles combined, contributing to roughly one in every five child deaths around the world. Presently, the major challenge lies in identifying the disease at preliminary stage so that its type can be identified, and appropriate precautions can be taken. Typically, a radiologist performs diagnosis to detect pneumonia by analysing the chest X-ray images of different patients. However, the number of such specialists are very less compared to the 450&#xa0;million people affected by pneumonia every year. This challenge can be addressed by incorporating Machine Learning models into pneumonia diagnosis. In this paper, a pneumonia detecting model is proposed that uses Image Processing and Deep Learning techniques employed on chest X-ray images. This AI-driven approach uses the Convolutional Neural Network (CNN) algorithm, particularly the VGG16 model, to extract deep features from the images for precise classification. This model will be trained and tested on an extensively used chest radiography dataset, aspiring to achieve high accuracy in distinguishing between normal and pneumonia-affected lungs. The results of the proposed model are quite impressive, with an outstanding performance of a precision of 92.73%, accuracy of 92.63%, recall of 95.90%, F1 score of 94.21%, and AUC of 0.9872 on the test data. The obtained results indicate that the model is highly effective and can be integrated into an automated Pneumonia diagnosis system, assisting in early detection and timely treatment.</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

PneumoNet: A VGG16 Based CNN Model for Pneumonia Prediction

  • 期刊论文

  • 发表日期：

    2025年12月24日

查看全文
  • Sanjit Kumar Dash,
  • Bimalendu Nanda,
  • Kanha Charan Dash,
  • Pradipta Kumar Mishra,
  • Satyajit Pattnaik

摘要

Pneumonia is a virulent disease that has caused deaths of millions of people globally. Each year, it claims more children’s lives than malaria, AIDS, and measles combined, contributing to roughly one in every five child deaths around the world. Presently, the major challenge lies in identifying the disease at preliminary stage so that its type can be identified, and appropriate precautions can be taken. Typically, a radiologist performs diagnosis to detect pneumonia by analysing the chest X-ray images of different patients. However, the number of such specialists are very less compared to the 450 million people affected by pneumonia every year. This challenge can be addressed by incorporating Machine Learning models into pneumonia diagnosis. In this paper, a pneumonia detecting model is proposed that uses Image Processing and Deep Learning techniques employed on chest X-ray images. This AI-driven approach uses the Convolutional Neural Network (CNN) algorithm, particularly the VGG16 model, to extract deep features from the images for precise classification. This model will be trained and tested on an extensively used chest radiography dataset, aspiring to achieve high accuracy in distinguishing between normal and pneumonia-affected lungs. The results of the proposed model are quite impressive, with an outstanding performance of a precision of 92.73%, accuracy of 92.63%, recall of 95.90%, F1 score of 94.21%, and AUC of 0.9872 on the test data. The obtained results indicate that the model is highly effective and can be integrated into an automated Pneumonia diagnosis system, assisting in early detection and timely treatment.