Application process & FAQ
Medical Imaging Data Scientist 80%
The Department of Ophthalmology at the Universitätsspital Zürich is an active contributor to the Fight Retinal Blindness! (FRB!) registry, an international, clinician-led initiative collecting real-world longitudinal data on retinal diseases.Retinal diseases such as AMD and DME are major causes of visual impairment worldwide. With the increasing availability of longitudinal ophthalmic imaging, especially Optical Coherence Tomography (OCT), and detailed clinical records, there is a growing opportunity to leverage AI to support and optimize personalized treatment decisions.
Despite rapid advances in medical AI, real-world clinical decision support systems for retinal diseases are only limited to unimodality and/or static predictions, and are limited by challenges such as data heterogeneity, temporal variability, and the integration of multimodal inputs (e.g., imaging and clinical data). Foundation models (large pretrained models for visual and tabular inputs) present a promising approach for robust feature extraction and multimodal learning in this domain.
This project aims to address these limitations by developing and evaluating multimodal, longitudinal AI models, leveraging foundation models for OCT imaging and clinical time-series data.
Your main tasks
You will contribute and test different parts in the development of AI models analysing longitudinal OCT B-scans.
Key objectives include:
Key objectives include:
- Integration and preprocessing of multimodal longitudinal data
- Fine-tuning OCT foundation models for disease progression modelling
- Designing multimodal fusion strategies (image + clinical data)
Your profile
Essential:
Desirable:
Soft Skills
- Master's degree in data science, biomedical image analysis or related
- Very strong background in AI for medical imaging (Pytorch)
- Previous experience with longitudinal / time-series modelling
- Solid understanding of transformer architectures, especially Vision Transformers (ViT) and attention-based multimodal models
- Good communication skills using lay language
- English proficiency
- Strong interest in the medical and clinical domain
- Motivation to understand disease mechanisms and clinical workflows
- Willingness to collaborate closely with clinicians
Desirable:
- Familiarity with multimodal foundation models
- German is a plus
Soft Skills
- Strong organization and experiment tracking skills
- Motivation to read and critically assess scientific literature
What we offer
- An international multidisciplinary team combining Ophthalmologists with deep clinical expertise, Statisticians, Data scientists and AI researchers' collaboration
- Regular supervision and technical guidance
- Exposure to cutting-edge research questions at the interface of AI and ophthalmology
- Opportunity to contribute to publishable research and future clinical tools
Application process and further information
Interested candidates should submit their Curriculum Vitae (CV) highlighting relevant experience and a Motivation Letter.
For further questions, please contact Camille Delgrange, Scientific Collaborator, via mail.
For further questions, please contact Camille Delgrange, Scientific Collaborator, via mail.
Your application
Stephanie IndlekoferRecruiting
Benefits
About USZ
The University Hospital of Zurich is one of the largest hospitals in Switzerland. Our more than 8,600 employees and around 700 trainees work every day for the well-being of our patients.
Education und Training
More about the faculty