Don't miss our weekly PhD newsletter | Sign up now Don't miss our weekly PhD newsletter | Sign up now

  Deep learning and bioinformatics approaches for personalised medicine

   School of Cancer and Pharmaceutical Sciences

This project is no longer listed on and may not be available.

Click here to search for PhD studentship opportunities
  Dr Heba Sailem  Applications accepted all year round  Self-Funded PhD Students Only

About the Project

An exciting opportunity to work within a highly multidisciplinary group at the Biomedical AI and Data Science Lab ( at the School of Cancer and Pharmaceutical Sciences to develop new computational pathology approaches to improve cancer patient treatment and diagnosis. This interdisciplinary project is suited for graduates with a background in Computer Sciences, Biomedical Engineering, Computer Vision, Bioinformatics, or Mathematics. Candidates with interest in lab-based projects would also be considered.

Project description

Breast cancer is the most common cancer in the UK. A major unmet need is the uncertainty in best treatment strategy in patients diagnosed at an advanced stage of the disease. For instance, the benefit of treating patients diagnosed with triple negative breast cancer remains highly variable. Moreover, preventive strategies that include surgery in patients susceptible to developing breast cancer, such as those diagnosed with ductal carcinoma, result in high economic burden, and greatly impact patient quality of life. Therefore, more individualised treatment strategies are urgently needed.

Histopathology remains the gold standard in cancer patient diagnosis and involves the examination of phenotypic and morphological changes in cells and tissues. It also allows evaluating the tumour heterogeneity and immune microenvironment as thousands of cells can be imaged per tumour along their spatial arrangement and colocalisation.

Artificial intelligence (AI) and machine learning approaches revolutionised our ability to process and discover patterns from large datasets. They also demonstrated great success in classifying large-scale cancer images and predicting patient diagnosis. However, a major challenge in translating AI models to the clinic is the difficulty of interpreting their performance and their limited generalisability. Predicting patient response to treatment can also be confounded by many factors such as variability in clinical practice and the used equipment.

In this project you will leverage advances in AI and deep neural networks to develop robust algorithms for predicting patient response to treatment based on systematic scoring of tumour composition and topology. You will interrogate these models using genomic and clinical data to interpret their biological relevance. You will have the opportunity to develop effective visualisation approaches of these data. You will work closely with clinicians to design bespoke explainability and visualisation approaches that ensure the clinical utilisation of these approaches. Such methods can achieve multiple advantages: 1) stratify patients who are more likely to benefit from a certain treatment 2) determine alternative treatment pathways for patients who are unlikely to benefit from the treatment which include enrolling them to participate in future clinical trials.

Over the tenure of this project you will work with renowned clinical experts and gain diverse skills in state-of-the-art computer vision, deep learning, computational and digital pathology techniques.

How to apply

Please email your CV and a personal statement describing how your previous experience and research interest fit the selected project to [Email Address Removed].

Biological Sciences (4) Computer Science (8) Engineering (12)

 About the Project