Project Overview:
In this project, our team aimed to develop a Convolutional Neural Network (CNN) capable of accurately classifying chest X-ray
imaging into two categories: Normal or Pneumonia. The dataset used for training and evaluation comprised of 5,863 chest X-ray
images obtained from pediatric patients of one to five years old. These images were sourced from the Guangzhou Women and
Children's Medical center in China.
Objective:
The primary objective of our project was to leverage deep learning techniques, specially CNNs, to create a reliable pneumonia
image classifier. By accurately identifying cases of pneumonia from chest X-ray images, our model could assist healthcare
professionals in diagnosing and treating pedriatic patients more efficiently. Early detection of pneumonia through automated
image analysis has the potential to improve patient outcomes and reduce the burden on medical resources, specially in places
where medical resources are scarce.
Approach:
1. Data Collection and Preprocessing
2. Model Development and Optimization
3. Model Training and Evaluation
Results:
Our CNN model achieved a high level of accuracy (~98%) in classifying chest X-ray images as Normal or Pneumonia. The model's
performance was validated through comprehensive evaluation metrics, demonstrating its potential utility in clinical settings.
Ethical Considerations:
All of the X-ray imaging was performed as part of the patients routine clinical care. The objective of this CNN model
is to provide possible improvement of accurate and timely diagnosis of pneumonia. The dataset should only be used with a
purpose similar to this.
Data Source:
https://www.kaggle.com/datasets/paultimothymooney/chest-xray-pneumonia
Relevant Links:
https://www.youtube.com/watch?v=jztwpsIzEGc&t=263s
https://www.cell.com/cell/fulltext/S0092-8674(18)30154-5