Postdoc in Advanced Reconstruction Algorithms for Time-Resolved Nano CT of Sustainable Cement

Would you like to put your mathematical and algorithm skills to work in a truly interdisciplinary project and be part of devising new sustainable cement formulations through improved dynamic X-ray imaging, then come to DTU and work with us!  

We are seeking a full-time postdoctoral researcher to join DTU Compute as part of the Villum Synergy funded project “MathCrete”. This research position will explore novel image reconstruction and segmentation algorithms for improving the contrast and space+time resolution of dynamic in situ micro and nano CT (computed tomography).

The position offers the unique opportunity to collaborate closely with mathematical imaging researchers at DTU Compute, the DTU 3D Imaging Center and cement experts at DTU Construct, in particular another MathCrete postdoc focused on developing new cement formulations and characterization techniques. For algorithms and software there will also be the opportunity to work with the team behind the Core Imaging Library (CIL) and contribute to open-source reproducible research.

Together, we aim to push the boundaries of mathematical imaging data science, materials science and digital building technology! 

The period of employment is 2 years. Tentative starting date: Summer 2025.

Responsibilities and qualifications
Your focus will be on developing, implementing and applying advanced image reconstruction algorithms to improve the quality of time-resolved micro and nano X-ray CT imaging of cement mixing processes. This is challenging due to the fast-evolving chemistry in the early mixing phase and low X-ray contrast between mixing and intermediary components. The goal is through improved time-resolved reconstruction to enable characterization of the cement microstructure in 3D + time to help understand and develop more sustainable cement formulations. This may involve techniques such as spatio-temporal regularization, discrete tomography, low-dimensional latent representations and machine learning.

The ultimate aim is to reduce the carbon footprint for the construction industry and enable CO2 capture from the atmosphere.

Your objectives will include to: 

• Develop new optimization and/or machine-learning based reconstruction and segmentation algorithms to improve image quality in time-resolved micro and nano CT data.
• Develop and apply numerical implementations in simulation studies as well as to large experimental micro and nano CT data.
• Explore multi-modality reconstruction techniques to incorporate additional characterization data such as SEM, TEM, RAMAN, XRD data.
• Collaborate closely with an interdisciplinary team at DTU Compute, the DTU 3D Imaging Center and DTU Construct, in particular the other “MathCrete” postdoc at DTU Construct.
• Publish results as journal articles and disseminate at scientific conferences.
• Possibility to co-supervise PhD/MSc/BSc students and contribute to teaching. 

Qualifications: 

• You should have (or be close to achieving) a PhD degree.
• Background within computational methods for inverse problems, ideally tomography.
• Experience with development of numerical implementations, ideally in Python.
• Experience with X-ray or neutron tomography data is a plus. 

Flexibility is essential. We are looking for a team player who can also work independently, who is motivated to help and be part of a dynamic research environment; and who has a positive, supportive and creative outlook. We expect you to have good communication and personal interaction skills and to be fluent in English.

As a formal qualification, you must hold a PhD degree (or equivalent).

Application procedure 
To apply, please read the full job advertisement via the 'Apply' button above.

Application deadline: 15 June 2025

Advert information