Computing Degree Show 2018

Machine Learning for Predicting Brain Tumour Mutation Status based on Multi-parametric MR Images

This project provides a software system that implements a multi-parametric brain image annotation tool in combination with an existing machine learning algorithm that uses annotated data and a prediction model to predict the brain tumour mutation status of an annotated brain image. By merging these two main functionalities together, the process of acquiring a quick prediction is streamlined significantly. The system is easy to use out-of-the-box and will be familiar to users with previous experience of medical image manipulation software, whilst also being intuitive and simple to use for less experienced users. The software system supports the analysis of multi-parametric DICOM brain images, and is implemented using the Java programming language, providing compatibility with a wide variety of operating systems.

Iain Morton