Radiology Assistant
A CNN-powered tool to classify chest X-rays as normal or abnormal using the NIH dataset.
Overview
This project implements a convolutional neural network (CNN) to classify chest X-rays as either normal or abnormal using the NIH Chest X-ray dataset from Kaggle.
It was developed as part of my computer science capstone, which required applying machine learning to address a real-world industry or business need. The binary classification requirement led to a model that outputs a straightforward normal vs. abnormal result.
Purpose
The goal of this project is to provide personalized movie recommendations that improve over time with user input. Recommendation engines are vital to modern digital platforms, from Netflix to Spotify, and this project mimics their fundamental mechanics for educational purposes.
It showcases my understanding of data manipulation, similarity algorithms, and model evaluation techniques in a real-world context.
Technologies
- Python
- TensorFlow
- Keras
- Scikit-learn
Features
- Compatible with both Google Colab and Jupyter Notebook
- Includes a demonstration script for testing