Project description german/english: Ausschreibung_Masterarbeit_MRT
Institution: ZAR Tübingen / Department of Neurology & Hertie Institute for Clinical Brain Research,
University Hospital Tübingen
Contact: Dr. med. Anne Lieb (anne.lieb@uni-tuebingen.de)
Start: By arrangement, immediately possible
Scope: Master’s thesis (approx. 6 months) or research internship with master’s thesis option; entry
into a PhD project possible upon interest
Language: German or English
The Project Working title: „Structural and Lesion-Network Signatures of Response to Contralesional Low-Frequency rTMS in Chronic Stroke“
Background
Since 2019, over 112 patients with chronic post-stroke arm paresis (upper limb plegia) have been treated at the TMS outpatient clinic of the University Hospital Tübingen using low frequency repetitive TMS (1 Hz, contralesional, neuronavigated). Clinical evaluation demonstrates significant reductions in spasticity and moderate motor improvements — consolidated in a manuscript. The open question addressed by the follow-up manuscript: Which structural characteristics of the individual stroke explain why some patients respond to stimulation — and others do not?
Individual structural MRI datasets are available for all 112 patients (MPRAGE 3 T + DTI), originally acquired for neuronavigation. These will now be systematically analysed — from lesion segmentation through DTI-based tractography to Lesion Network Mapping following Fox (a methodologically state
of-the-art technique that embeds individual lesions into a normative functional connectome and
identifies responder networks).
Tasks
- Lesion segmentation (semi-automated + manual quality control, n = 112; tools: ITK-SNAP, LINDA/LST, SynthSeg/FastSurfer)
- Preprocessing & Normalisation into MNI standard space (ANTs, FSL, lesion-aware registration)
- DTI analysis of the corticospinal tract: Tractography (FSL probtrackx / MRtrix3), FA asymmetry calculation, CST lesion load according to SMATT atlas
- Lesion Network Mapping (Lead-DBS / normative GSP connectome, n = 1,000 healthy
subjects): individual connectivity maps, voxel-wise group comparison (FSL randomise, TFCE
correction) - Statistical analysis: responders vs. non-responders, covariate adjustment, cross-validation
- Contribution to manuscript preparation
What we offer
- Fully prepared clinical data matrix: n = 112 patients, FMA-UE / Modified Ashworth Scale
pre/post, MEP data, responder labels based on MCID criteria - Structured analysis plan (step model 1.1–1.4), ready for immediate entry
- Positive ethics approval (reference number 2026-0203), patient consent forms available
- Insight into the full clinical context of rTMS neurorehabilitation at the Centre for Outpatient
Rehabilitation Tübingen - Possible continuation as PhD project upon mutual interest
What we are looking for
- Studies in Cognitive Science, Computational Neuroscience, Neuroinformatics, Medical Physics, Computer Science with a focus on medical image processing, or related disciplines
- Basic knowledge of MRI image processing (FSL, FreeSurfer, SPM, ANTs or equivalent) or strong
Python/MATLAB skills and ability to learn quickly - Interest in clinical neuroscience and the mechanisms of brain reorganisation after stroke
- Careful and independent work — quality control of segmentations contributes directly to publication quality
Prior experience with diffusion MRI, connectivity analyses, or lesion segmentation tools would be ideal, but is not essential — methodological interest and willingness to learn are what matter most.
Contact
Dr. med. Anne Lieb
Hertie Institute for Clinical Brain Research, Department of Neurology and Centre for Outpatient Rehabilitation Tübingen University Hospital Tübingen
✉ anne.lieb@uni-tuebingen.de
Please send a brief statement of motivation and your CV. An informal introductory conversation by email or Teams is welcome.