MASTER Assignment
Artificial intelligence against cancer
Type : Master CS
Location: University of Twente
Period: t.b.d.
Student: (Unassigned)
If you are interested please contact:
Background:
This project is about helping medical doctors to improve the treatment of patients with cancer. It will involve using machine learning techniques to build Bayesian-network models that give insight into the factors to are crucial in improving the survival of patiens after treatment (surgery, radiotherapy, and chemotherapy). In this project the focus will be on investigating international data from patients with endometrial cancer, i.e. cancer of the womb. You need to have some affinity with the clinical domain, and the project will be carried out in close collaboation with the clinicians at RadboudUMC in Nijmegen. However, the actual work will be done at UTwente and it is not necessary to have medical knowledge before you start.
Assignment:
The Department of Gynecology and Obstetrics of RaboudUMC (the big academic hospital of Raboud University) are currently investigating in collaboration with the European Network of Individual Treatment in Endometrial Cancer (ENITEC, https://www.esgo.org/network/enitec/) the best way to improve and personalise the treatment of endometrial cancer. For this purpose, large datasets with data from patients from some of the EU countries, have been made available. The data incorporates many variables, from personal characteristics, such as age, to clinical signs and symptoms, results from radiological investigations, pathology, and various specialised and common laboratory tests, treatments, and outcome after treatment.
The purpose of this master project is to develop a causal Bayesian network that helps both clinicians and patients in deciding about the best possible clinical management for the individual patient. Given the need to be able to understand the recommendations provided by the model, we opt for a model-based approach, using Bayesian networks. The research involves Bayesian network structure learning and validation.
Profile of the student:
You need to have an interest in clinical medicine as an application field and be willing to interact with medical doctors. Knowledge of machine learning and AI is also required.