MRI Data Augmentation Using GANs (MTHE 493 Thesis)

Summary

The use of machine learning for healthcare has a lot of potential uses, one of them being faster and more accurate diagnosis. Medical data collection with the sole use of building datasets is too expensive which often leaves researchers with small datasets. In addition to being small, they often have isses with class imbalence and do not reflect real world data. This project uses Generative Adversarial Networks (GANs) to address the issues of data scarcity and class imbalence by exploring the use of synthetic data.

Contributions

My primary contribution to this project involved building the GAN using the KERAS Python library, training the models, implementing novel loss functions, and writing sections of the reports.