Detecting Pneumonia Through Chest X-Ray Analysis
Home → Research → Detecting Pneumonia Through Chest X-Ray Analysis
RESEARCH PUBLICATION · DEEP LEARNING · MEDICAL IMAGE ANALYSIS
This research explores the use of convolutional neural networks (CNNs) to classify chest X-ray images for pneumonia detection. The work investigates a deep-learning-based image classification workflow and examines model performance using validation results and visualizations.
A conference paper examining deep-learning-based chest X-ray classification and model evaluation.
Publication type
Conference paper
Conference series
ICAIN 2024 · Springer Nature
Online publication
1 October 2025
Research Problem
Chest X-ray interpretation can be challenging because pneumonia-related patterns may be subtle or overlap with other lung conditions. The paper explores how deep-learning-based image classification can support the analysis of chest X-rays.
Methodology
The study performs chest X-ray image classification across the dataset’s classes using a convolutional neural network (CNN) architecture. Stochastic Gradient Descent (SGD) is used for optimization with Sparse Categorical Cross-Entropy as the loss function. Accuracy is used as an evaluation metric, alongside visualizations of misclassified images and training and validation performance curves.
Key Findings
The paper reports a deep-learning workflow for classifying chest X-ray images and reviewing its validation behavior through metrics and visualizations. The findings are presented as research evaluation of the described approach; they do not establish clinical validation, diagnostic readiness, or clinical outcomes.
Publication Details
Title: Detecting Pneumonia Through Chest X-Ray Analysis
Authors: Vedaant Melkari, Harsh Motiramani, Malav Mehta, Archana Nanade, and Abhay Kolhe
Publication type: Conference paper
Conference series: International Conference on Artificial Intelligence and Networking (ICAIN 2024)
Publisher: Springer Nature
Series: Lecture Notes in Networks and Systems, Volume 1269
Pages: 461–476
Online publication date: 1 October 2025