PhD Position in Advanced Brain PET Image Reconstruction
The Molecular Image Reconstruction Group at KU Leuven (Department of Imaging & Pathology), led by Prof. Georg Schramm, conducts research at the intersection of medical image reconstruction, positron emission tomography (PET), and machine learning. The group has close ties with UZ Leuven, one of Europe’s leading academic hospitals, providing direct access to state‑of‑the‑art clinical PET infrastructure and expertise in nuclear medicine and neurology.
This position is embedded in the PEARL project (PET Enhancement via Advanced Reconstruction using variational methods and machine learning), a tri‑national Weave project funded by FWO (Belgium), FWF (Austria), and DFG (Germany), in collaboration with the University of Graz and TU Munich.
Responsibilities
* Data curation: retrospective collection, quality control, and preprocessing of approximately 500 static and dynamic brain PET/MR datasets acquired at UZ Leuven during the last 10 years.
* Open data infrastructure: pseudonymisation and conversion of raw list‑mode data into open, standardized formats.
* GPU software development: design and optimisation of high‑performance, open‑source time‑of‑flight (TOF) list‑mode forward and back‑projectors for integration into machine‑learning frameworks such as PyTorch, building on the existing parallelproj library.
* Method evaluation: critical assessment of reconstruction algorithms developed across all partner institutions, in close collaboration with clinical and kinetic‑modelling experts at UZ Leuven.
* Challenge organisation: planning and execution of an open international brain PET reconstruction challenge, including data‑sharing agreements, automatic evaluation pipelines, and platform setup.
* Dissemination: presentation of results at international conferences and publication in peer‑reviewed journals.
Profile
* Master’s degree (or equivalent) in Engineering, Physics, Computer Science, Mathematics, or a closely related field, obtained before the start date.
* Strong programming skills in Python (or other high‑level languages) and experience with scientific computing libraries (e.g. NumPy, PyTorch).
* Strong mathematical background in linear algebra, optimisation, and probability/statistics.
* Very good written and spoken English.
* Enthusiasm for interdisciplinary research at the interface of medical imaging and machine learning.
* High intrinsic motivation.
* Prior experience with medical image reconstruction or emission tomography.
* Experience in efficient handling of large research data sets.
* Knowledge of variational methods, convex optimisation, or iterative algorithms for inverse problems.
* Experience with open‑source software development and version control (Git).
* Exposure to deep‑learning methods.
Offer
* A fully funded 4‑year PhD position within a stimulating, internationally connected research environment.
* Direct access to state‑of‑the‑art clinical PET infrastructure and clinical experts at UZ Leuven.
* High‑performance GPU computing resources.
* Participation in international conferences and project workshops (Leuven, Munich, Graz).
* Short‑term research exchanges at partner institutions in Graz (Austria) and Munich (Germany).
* A vibrant, international academic community in Leuven – a historic university city in the heart of Europe.
Contact
For more information please contact Prof. Georg Schramm, mail: georg.schramm@kuleuven.be.
EEO Statement
KU Leuven strives for an inclusive, respectful and socially safe environment. We embrace diversity among individuals and groups as an asset. Open dialogue and differences in perspective are essential for an ambitious research and educational environment. In our commitment to equal opportunity, we recognize the consequences of historical inequalities. We do not accept any form of discrimination based on, but not limited to, gender identity and expression, sexual orientation, age, ethnic or national background, skin colour, religious and philosophical diversity, neurodivergence, employment disability, health, or socioeconomic status. For questions about accessibility or support offered, we are happy to assist you at this email address.
#J-18808-Ljbffr