Logo Lanfrica

Ayush7921/Pneumonia-Detection-AI-ML

Domain:

healthcare

Record type:

software
Creator:
Ayu
Host:
🫁⚑ A lightweight machine learning web app that detects Pneumonia from chest X-rays with 94.6% accuracy in under 1.5 seconds. Built for low-resource clinics. # 🫁 Pneumonia Detection System using Chest X-Rays through Machine Learning Project website: PNEUMONIA DETECTION An intelligent web-based application that detects pneumonia from chest X-ray images using classical Machine Learning techniques. The system combines image preprocessing, handcrafted feature extraction, ensemble classification, and explainable AI to provide accurate, fast, and interpretable predictions. --- ## πŸ“– Overview Pneumonia is one of the leading causes of respiratory-related deaths worldwide, especially among children and elderly patients. Accurate diagnosis from chest X-rays requires experienced radiologists, which may not always be available in rural or resource-constrained healthcare facilities. This project provides an AI-assisted diagnosis system that analyzes chest X-ray images and predicts whether a patient is suffering from pneumonia. Unlike many deep learning-only approaches, this project focuses on an interpretable and lightweight machine learning pipeline that can be deployed efficiently without requiring high-end GPU infrastructure. --- ## ✨ Key Features - 🩻 Chest X-ray image upload - βœ… Four-layer X-ray validation system - 🎨 CLAHE-based image enhancement - πŸ“Š HOG (Histogram of Oriented Gradients) feature extraction - 🧩 LBP (Local Binary Pattern) feature extraction - βš–οΈ Feature normalization using StandardScaler - πŸ€– Ensemble Machine Learning classifier - πŸ“ˆ Pneumonia severity scoring - πŸ” SHAP-based explainability - πŸ“„ PDF report generation - 🌐 FastAPI backend - πŸ’» Responsive web interface - ☁️ Serverless deployment --- # πŸ—οΈ System Workflow ``` Chest X-ray Upload β”‚ β–Ό 4-Layer Image Validation β”‚ β–Ό Image Preprocessing (CLAHE) β”‚ β–Ό Feature Extraction (HOG + LBP + Statistical Features) β”‚ β–Ό Feature Scaling (StandardScaler) β”‚ β–Ό Ensemble Classifier (SVM + Random Forest + Gradient Boosting) β”‚ β–Ό Prediction β”‚ β–Ό Severity Analysis β”‚ β–Ό SHAP Explainability β”‚ β–Ό Result & PDF Report ``` --- # 🧠 Machine Learning Pipeline ### Image Preprocessing - …