Deep learning has revolutionised the field of medical imaging analysis by automating tasks such as Anomaly Detection and image segmentation. Despite its success, the deployment of Deep Learning models in clinical practice is limited by their requirement for large annotated datasets. The process of obtaining annotations requires specific domain expertise which is both tedious and costly. This project aims to address this challenge by exploring Few Shot techniques for Anomaly Detection and segmentation in order to reduce the barriers to clinical implementation of Deep Learning models.