2023-24-project-catalogue

###Personalised radiotherapy using multi - modal artificial intelligence for improved lung cancer radiotherapy

Project ID: 2228bd1205 (You will need this ID for your application)

UCL Lead department: Medical Physics and Biomedical Engineering

Lead Supervisor: Charles - Antoine Collins Fekete

Project Summary:

Background: The student will help accelerate the development of outcome prediction and personalised radiotherapy for lung cancer, through the analysis of anatomical, functional, biological and clinical information with state-of-the art artificial intelligence. They will have access to a large database (>2000) of pre-scanned lung cancer patient of treated with concurrent chemo-radiotherapy, ready for analysis. This project involve data discovery, data analysis, computational modelling through artificial intelligence, outcome prediction and experimental validation in partnership with our industrial (e.g. Microsoft) and clinical collaborators (e.g. UCLH). Project: The overall goal of the project is to develop a personalised therapy plan for patients suffering from non-small cell lung cancer (NSCLC) based on the outcome predicted from artificial intelligence algorithms trained on a large multi-modal (radiological, biological images) dataset of patients. Inoperable NSCLC is a severe disease for which the actual chemo-radiotherapy treatment has remained mostly unchanged for more than 30 years with a poor 18.6% 5-year survival. This survival is limited by the diverse spectrum of clinical presentation within the NSCLC patient’s group which limits the one-solution-for-all current approach to treatment cancer. Artificial intelligence is particularly successful at extracting predictive feature from a large and varying database of images and therefore highly suited for this type of study. Objectives:

To predict personalised NSCLC outcomes (2 year survival, lung and oesophagus toxicity) using radiological input data in classification neural network (1st Year).
To introduce biological features in the outcome prediction model through scanned histopathology slides (2nd Year).
To integrate the predicted AI outcomes and multi-modal input in a cost-function towards a personalised radiotherapy treatment plan (3rd Year). Research Environment: The project is embedded within the Radiotherapy Research group at UCL. The available infrastructure include access to large scale GPU clusters and clinical/AI expertise.