Back to projects
Tuberculosis Detection from X-Rays
CNN trained with transfer learning to classify tuberculosis in 630 chest X-ray images, evaluated with accuracy, precision, recall and F1.
- Project
- MSc AI and Machine Vision module
- Role
- Computer vision
- Period
- 2026
[ Overview ]
The problem
Screening chest X-rays for tuberculosis needs accurate classification from a small labelled dataset.
What I built
Fine-tuned a pretrained convolutional neural network on 630 chest X-rays and evaluated it with standard classification metrics.
[ Highlights ]
- Convolutional neural network trained with transfer learning
- 630 labelled chest X-ray images
- Evaluated with accuracy, precision, recall and F1
[ Stack ]
- Python
- CNN
- Transfer learning
- Computer vision