Deep Learning in Medical Imaging Analysis

A site discussing the uses of AI in Radiology

January 2018 - March 2018

This project was undertaken as my first year group project, and involved creating a website discussing the uses of AI in Medical Imaging Analysis, aimed at a specialist.

Our main areas of research were to look into the history and evolution of medical imaging and machine learning as separate topics, and then find areas in which they overlapped. We looked at Convolutional Neural Networks, along with Semantic Segmentation, as these are the most commonplace in medical imaging. We then looked at overlapping topics, these being the uses of AI in radiology, and a deeper case study into the user for detection of lung cancer.

Getting to know all of this information about AI was great, but we also had to present it in the form of a website. We had to create the site from scratch (without using a framework, such as Wordpress). We wrote the in Markdown, and used a Bash script (which invoked Pandoc) compiled it to HTML. Since we didn't use a web framework to write the CSS, it increased the challenge, but also meant that after the project, we had a more thorough understanding of CSS. Mobile optimisation was one particular challenge we found difficult.

We won the prize for the best project in our category of AI in Medical Imaging, and second best project overall.