You cannot apply for this job anymore (deadline was 20 Feb 2023).
Browse the current job offers or choose an item in the top navigation above.
Sharing clinical data and therefore collecting large-scale representative multi-center datasets is difficult, or sometimes even impossible, because of privacy, safety and regulatory issues. The research field of federated learning addresses this problem by training algorithms collaboratively without exchanging the data itself. This project aims to study this technique for improving etiological diagnosis of dementia. Accurate etiological diagnosis of dementia is crucial for care and treatment decisions, but current clinical criteria fail to differentiate etiologies due to overlap in early-stage symptoms. Machine learning methods can support clinicians in this decision making by combining all available information from previous patients for the diagnosis of current patients. While such an approach is highly promising for etiological dementia diagnosis, large-scale representative multi-center data as well as interdisciplinary collaboration are crucial to take the next step.
This project aims to investigate federated technology for development and validation of AI using multi-center clinical and radiological data, focusing on improvement of etiological diagnosis of dementia.As PhD candidate on this project you will be part of a large national consortium (TAP-Dementia), aimed at improving diagnostics of Alzheimer's disease and non-Alzheimer's dementia.This consortium provides access to multi-center data as well as knowledge from multiple disciplines. In this PhD project, you will study state-of-the-art solutions for federated learning as well as develop and validate machine learning and deep learning techniques for etiological diagnosis of dementia using federated imaging and clinical data. In addition, you will closely collaborate with researchers and clinicians from several Alzheimer centers in the Netherlands to implement a federated data network and to establish a federated validation framework for diagnosis of dementia.
Erasmus MC (University Medical Center Rotterdam)
- You should be an independent and highly motivated researcher with an MSc or MEng degree in a technical domain (e.g., Computer Science, Physics, Engineering, Mathematics, Clinical/ Biomedical Technology).
- Excellent proficiency in English is required.
- You should be able to communicate and work with researchers from various fields.
- Experience with machine learning and programming is highly valued.
Conditions of employment
- You will receive a temporary position for 4 years. The gross monthly salary is € 2.789 ,- in the 1st year and increases to € 3.536,- in the 4th year (scale OIO).
- Excellent fringe benefits, such as a 13th month that is already paid out in November and an individual travel expense package.
- Pension insurance with ABP, we take care of approximately 2/3 of the monthly contribution.
- Special benefits, such as an incompany physiotherapist and bicycle repairer. There is also a gym where you can work on your fitness after work.
- An international office which aids you in preparing for your arrival and stay.
The research will be conducted at the Biomedical Imaging Group Rotterdam (BIGR), which is part of the Department of Radiology & Nuclear Medicine of the Erasmus MC, University Medical Center Rotterdam. The department has an international character with a good balance between internationally recognised, high-end research and an excellent social working environment. The supervision for this project will be provided by Dr. Esther Bron, Assistant Professor in Neuroimage Analysis and Machine Learning, and Dr. Ir. Stefan Klein, Associate Professor in Medical Image Analysis and Machine Learning. You will collaborate with a team of experienced researchers with backgrounds in image analysis, machine learning, radiology, neurology and neuropsychology.