- Department/Institute:
- Clinic and Polyclinic for Nuclear Medicine / Nuclear Immunology Lab
- Subject area:
- AI in Biomedical Imaging and Translational Neuroscience
- Name of supervisor:
- Prof. Dr. Dr. Matthias Brendel
- Number of open position:
- 1
- Project title:
- Multimodal Image Integration Pipeline for Neurodegeneration, Tumor and Immune Mapping in Rodents
- Project time:
- Full Doctoral Study-Model: 48 months
- Language requirements:
- Fluent command of English in speaking and writing
- Academic requirements:
- Candidates with a background in biomedical engineering, computer science, physics or neuroscience are encouraged to apply.
Experience with image processing techniques and proficiency in programming (Python, PyTorch/Keras/TensorFlow) are required.
Familiarity with deep learning microscopy or small animal imaging (PET/CT, MRI or light-sheet microscopy) would be advantageous.
Prior experience in quantitative image analysis, image registration techniques or multimodal data integration is desirable but not mandatory.
Project description:
Understanding disease mechanisms in the brain and body requires integrating imaging modalities that span different scales and contrasts. This project aims to develop a modular, multimodal image analysis pipeline that connects in vivo molecular imaging (PET/CT) with ex vivo high-resolution light-sheet microscopy for co-segmentation, registration, and quantification of disease markers in rodent models.
The initial focus will be on glioblastoma-bearing mouse brains, leveraging existing datasets with molecular labeling and tissue clearing. New datasets will be generated to expand model training and validation for supervised machine learning. Future applications may include models of neurodegeneration and systemic immune activation, as well as imaging of other organs (e.g., skull, heart).
On the computational side, the project will explore deep learning approaches for multimodal registration and segmentation, benchmarking single unified versus modular models, and evaluating their performance against established tools. The long-term goal is to create a reproducible research software product for automated multimodal image analysis in preclinical imaging.
The project combines computational modeling, image analysis, and experimental work in small-animal imaging, offering interdisciplinary training at the interface of AI, molecular imaging, and neuroimmunology.
To applicants: Please send following initial application documents to LMU-CSC Office before 15th December:
- Resume and Research Motivation Letter
- Certificate of Proficiency in English, equivalent to IELTS Test Academic 6.5 (no module below 6) or TOEFL IBT 95, is required
- Two letters of recommendation directly sent from your current Supervisors/Professors to LMU-CSC Office