Intracranial Hemorrhage Detection Using CNN with Attention

This project replicates a 2024 research paper that proposes a deep learning-based method to automatically detect intracranial hemorrhage (ICH) in CT brain scans. The goal is to learn how to read and implement research ideas using convolutional neural networks (CNNs) and attention mechanisms.

Highlights

• Achieved over 99% accuracy on public CT scan data
• Implemented CNN model with an attention mechanism
• Learned preprocessing, model building, training, and evaluation
• Based on the paper: "Automated Intracranial Hemorrhage Detection Using Deep Learning in Medical Image Analysis" (IEEE 2024)

This project is for educational and skill-building purposes.

---

Table of Contents

• Dataset
• Model Architecture
• Training & Results
• How to Run
• What I Learned
• Citation