Tuberculosis Detection from Chest X-Ray using Sparse CNN

📌 Overview

This project is a Computer-Aided Diagnosis (CAD) system that detects Tuberculosis (TB) from chest X-ray images.
It combines:

• Convolutional Neural Networks (CNNs) for deep feature extraction.
• Gray Level Co-occurrence Matrix (GLCM) for texture analysis.
• A Graphical User Interface (GUI) built with PyQt5 & Tkinter for easy use.

🎯 Objectives

• Automate TB detection in chest X-rays.
• Provide accurate classification into Normal or TB Present.
• Assist radiologists with quick and consistent screening.

🖥 Features

• Preprocessing of X-rays (resizing, grayscale, median filtering).
• Feature extraction using GLCM and CNN.
• Classification using a hybrid model.
• GUI for selecting images and displaying results.
• Confusion matrix visualization for model performance.

🛠 Technologies Used

• Python 3
• PyQt5, Tkinter, EasyGUI – for GUI
• OpenCV, scikit-image, PIL – for image processing
• TensorFlow/Keras – CNN model training
• Matplotlib, Seaborn – data visualization

📊 Example Output

Result Message: TB Present / Normal

Confusion Matrix: Model accuracy & misclassification rates

📌 Dataset

Public TB X-ray datasets used:

Montgomery County X-ray Set

Shenzhen Hospital X-ray Set

👩‍💻 Authors

Aditi Amol Londhe

Komal Kiran Mulik

Vaishnavi Vitthal Jadhav

Guided by Dr. Asma Shaikh (Annasaheb Dange College of Engineering & Technology, Maharashtra)