Deep Learning Project: COVID-19 Classification from Chest X-Rays

This project focuses on developing a two-layer deep learning model for automated classification of chest X-ray images into three categories: Normal, COVID-19, and Viral Pneumonia. By leveraging a two-step classification approach and convolutional neural networks (CNNs), the model achieves high accuracy in diagnosing COVID-19, aiding in early detection and intervention.

Project Overview

Problem Statement

The goal of the project is to build a deep learning model capable of: 1. Classifying images as Normal or Abnormal. 2. Further classifying abnormal images into COVID-19 or Viral Pneumonia.

Approach

The solution consists of two classifiers: • Classifier 1: Identifies whether an image is “Normal” or “Abnormal”. • Classifier 2: Further classifies “Abnormal” images into “COVID-19” or “Viral Pneumonia”.

Dataset • Classes: Normal, COVID-19, Viral Pneumonia • Dataset consists of grayscale X-ray images with varying levels of abnormality. • Data preprocessing includes normalization and resizing to fit the CNN input.

Proposed Architecture

Both classifiers use a similar CNN architecture:

1. Convolutional Layer: • 32 filters of size 3 * 3. • Activation: ReLU
2. Pooling Layer: • Max-pooling with a pool size of 2 * 2.
3. Flattening Layer: • Converts 2D feature maps into a 1D array.
4. Dropout: • Dropout rate of 0.3 to prevent overfitting.
5. Dense Layers: • Fully connected layer with 32 units (ReLU activation). • Output layer with 1 unit (Sigmoid activation for binary classification).

Conclusion

1. The two-layer CNN architecture effectively classifies chest X-ray images into Normal, COVID-19, and Viral Pneumonia with high accuracy.
2. Classifier 1 Accuracy: ~91%
3. Classifier 2 Accuracy: ~96%
4. The model demonstrates the potential for deep learning to assist in early diagnosis and treatment of COVID-19.