PhD fellowship in Biostatistics at the Department of Public Health


University of Copenhagen


September 1, 2023

We are offering a PhD fellowship with a special focus on statistical methods for image data collected in clinical trials on 1 November 2023 or as soon as possible hereafter. The PhD student is expected to develop and apply statistical methodology for causal analysis of observational and experimental data with focus on interpretable target parameters and rigorous statistical inference.

Project description

The PhD project aims at developing a statistical framework for studying high-dimensional mediators measured with medical imaging. Mediation analysis enables to study the role of variable on the causal pathway between a treatment and an outcome variable, and is thus of great interest to understand treatment mechanisms. While the case of a single mediator is well understood, difficulties arise with multiple mediators, especially in the high-dimensional case, as the number of possible paths going through the mediators increases super exponentially. A key part of this project is the integration of causal inference technics, elements from semi-parametric theory, and state-of-the-art machine learning techniques to develop a flexible yet interpretable approach. In particular, it should be able to account for dependencies between mediators, define mediation effects using counterfactuals, and have explicit assumptions under which a causal interpretation holds. The proposed framework will be applied on experimental data, e.g. to assess how the serotonin system mediates the effects of anti-depressive treatment in patients with major depressive disorders.

The project is a firmly based at Dep. of Public Health, University of Copenhagen but involves substantial collaboration with Peking University and Novo Nordisk. One or two extended research stays in Beijing are to be expected. The main goal of this larger collaboration is to develop new methodology and improve best practices in the analysis of image data in relation to clinical trials.

Principal supervisor is Professor Theis Lange, Section of Biostatistics,, Principal co-supervisor is Assistant Professor Brice Ozenne, Section of Biostatistics,

Job description

Your key tasks as a PhD student at SUND are:

  • Carrying through an independent research project under supervision.

  • Completing PhD courses or other equivalent education corresponding to approximately 30 ECTS points.

  • Participating in active research environments including a stay at another research team.

  • Obtaining experience with teaching or other types of dissemination related to your PhD project

  • Teaching and disseminating your knowledge.

  • Writing a PhD thesis on the grounds of your project

Key criteria for the assessment of applicants

The successful applicant for the Ph.D. scholarship will be a statistician with a strong background in mathematics and statistics i.e., a candidate degree in for example mathematics or statistics. Accordingly, we require documented skills within theoretical and applied statistics as well as high-level programming. Also, the applicant should have the ability to communicate research findings in teaching, conference talks and by writing scientific papers for international journals.

Please see the full advertisement and link to apply here: