Development of Novel Computational Intelligence Approaches to Cancer Prognosis and Diagnosis

An important challenge for primary and secondary care physicians remains the accurate evaluation of the risk of cancer occurrence and prediction of progression, both of which are essential for determining the optimum treatment and management. Predictive modelling in medicine involves deriving a mathematical model for the prediction of a future outcome for patients. Predictive tools can help during the complex decision-making processes, and provide individualised, evidence-based estimates for cancer patients. The predictive models can be based on statistical or computational intelligence techniques. Computational intelligence is a relatively new term, for which there is currently no formal definition. Computational intelligence algorithms are considered by some researchers to involve only evolutionary algorithms, neural networks, fuzzy logic, or hybrids of these. However, others consider a more broad definition of computational intelligence to include the above mentioned, as well as paradigms such as Bayesian belief networks, multi-agent systems, case-based reasoning and so on.

Many computational intelligence approaches, such as artificial neural networks and support vector machines, are known to increase accuracy in cancer prognosis and diagnosis because these approaches are capable of dealing with the complexity which is typically found in clinical datasets. However, a majority of these approaches do not provide qualitative reasoning behind the derived prediction.

The aim of the project is to identify the factors which are predictors of cancer prognosis and diagnosis tasks (prostate cancer and lung cancer), and to develop a new computational intelligence approach which can be used for cancer prediction, and which can achieve higher predictive accuracy than the existing approaches. The new computational intelligence method should be able to provide qualitative reasoning behind the prediction, and therefore provide the combinations of factors and their weightings which have derived the prediction. The derived method will be integrated in mobile apps and tablets. These apps will be capable of collecting and storing ‘big data’ which could be used for making cancer risk predictions in the future. Such data could include ‘conventional’ information on diet, physical activity, smoking, exposure to pollution, and aspirin use, and be enhanced on the basis of additional biomarkers such as peripheral blood phenotypic and other ‘liquid biopsy’ derived data. It is expected that the new approaches that will be developed will improve existing approaches for predicting cancer risk and the delivery of personalized approaches for the management and treatment of patients with cancer that are focused on reducing risk and recurrence.

Students will work with a subset of data extracted from The Health Improvement Network (THIN) database which is a large UK primary care database; and with data from the British Association of Urological Surgeons (BAUS).

For informal discussion regarding the project, please contact:
[Email Address Removed]

HOW TO APPLY
For further details please see the web site here:
http://www.ntu.ac.uk/research/graduate_school/studentships/132586.html

Please find attached an Application Form, notes for completion and guidance, and further details about the Schools and the available research projects.

*Applications from non-EU students are welcome, but a successful non-EU candidate would be responsible for paying the difference between non-EU and UK/EU fees. (Fees for 2015/16 are £12,300 for non-EU students and £4052 for UK/EU students)

ELIGIBILITY & ENTRY REQUIREMENTS
To be eligible to apply, you must hold, or expect to obtain by 1st October 2016, a Master’s degree, or a 1stClass/2.1 Bachelor’s degree in a relevant subject area (including, where appropriate, training in the relevant research methods and, where relevant, laboratory experience).

Please note that these scholarships are only available for new applicants. Existing PhD students are not eligible to apply


ENGLISH LANGUAGE REQUIREMENTS
Applications can be accepted from UK/EU and International students. The minimum English language proficiency requirement for candidates who have not undertaken a higher degree at a UK HE institution is IELTS 6.5 (with a minimum of 6.0 in all skills).