Multidimensional feature extraction of quantitative relaxation and diffusion brain MRI based on a cohort of ~ 120 subject datasets

Master thesis project

Project description

Magnetic resonance imaging (MRI) as we know it today is mainly assessed in a qualitative way, restricting image interpretation to features like size/volume or distance between objects [1]. However, tissue characterization can be supported by substantially more information sensitizing the MR system to physiological, biophysical or biochemical processes (e.g. relaxation, diffusion, perfusion etc.) [2]. Within this Master thesis project, we will retrospectively use MR data obtained in a follow-up study of a recent published work that focused on the investigation of learning-related changes in brain structural connectivity using diffusion MRI [3]. Within this study, about 120 healthy volunteers were measured at three different time points (baseline measurement, ~90min, after 24h) to assess learning-induced brain changes including the acquisition of phase-cycled balanced steady state free precession (bSSFP) data and standard diffusion tensor imaging (DTI) data. The bSSFP sequence is known to have a mixed sensitivity to the relaxation metrics of T1 and T2 [4] but also to microstructural tissue compartments with interesting signal asymmetries in white matter fiber tracts, possibly reflecting myelin [5,6]. Due to the very high SNR, speed and motion robustness, the bSSFP sequence is already used in a variety of approaches for fMRI [7,8] and relaxometry (simultaneous estimation of T1 and T2) [9,10]. Moreover, the observed signal asymmetries were explored for the estimation of diffusion metrics from bSSFP data using a neural network approach [11]. To bring all modalities together, reproducible and efficient processing pipelines will be required. This could be achieved with open-source frameworks like Nipype [12]. The goal of this Master thesis project is to process and combine the existing modalities. Focus will be on the processing of the bSSFP data and the analysis of the obtained quantitative relaxometry and asymmetry features. We aim at finding correlation of those features with DTI metrics. Depending on the results and progress, there will be the opportunity to explore new relationships between those features and the learning-induced changes in functional and diffusion MRI. The rough timeline for this project is structured as follows:


  1.  Familiarize yourself with the subject matter, data and processing. Get an understanding about MRI sequences, especially bSSFP and DTI but also fMRI, and to test the existing Nipype pipeline validated on 1-2 subjects. (~1-2 months) 
  2. Validate and improve the existing pipeline for data processing of >120 subjects. (~1-2 months)
  3. Data analysis (Statistics and Feature Extraction) (~2-3 months)
  4. Project summary (Presentation and Writing) (~1 month)

What we offer

  • Highly professional infrastructure to learn about MRI, Data processing, Neuroanatomy.
  • Great working environment with own workstation
  • Supervision by PhDs/PostDocs for this project

What we are looking for

  • Highly motivated graduate student with background in medical imaging, biotechnology, neuroinformatics, informatics or similar.
  • Computer skills: FSL, SPM (preferred/not required), programming (Python preferred (basic/advanced))
  • Great team work and communication skills
  • Available time of 6-8 Months.


Abramson RG, Su P-F, Shyr Y.
Quantitative metrics in clinical radiology reporting: a snapshot perspective from a single mixed academic-community practice.
Magnetic Resonance Imaging 2012; 30:1357–1366.
Seiberlich N, Gulani V (eds.)
Quantitative magnetic resonance imaging.
1st ed. Cambridge: Elsevier; 2020.
Brodt S, Gais S, Beck J, Erb M, Scheffler K, Schönauer M.
Fast track to the neocortex: A memory engram in the posterior parietal cortex.
2018; 362:1045–1048.
Bieri O, Scheffler K.
Fundamentals of balanced steady state free precession MRI: Fundamentals of Balanced SSFP MRI.
Journal of Magnetic Resonance Imaging 2013; 38:2–11.
Miller KL.
Asymmetries of the balanced SSFP profile. Part I: Theory and observation.
Magnetic Resonance in Medicine 2010; 63:385–395.
Miller KL, Smith SM, Jezzard P.
Asymmetries of the balanced SSFP profile. Part II: White matter.
Magnetic Resonance in Medicine 2010; 63:396–406.
Scheffler K, Ehses P.
High-resolution mapping of neuronal activation with balanced SSFP at 9.4 tesla: High-Resolution bSSFP Functional Imaging at 9.4T.
Magnetic Resonance in Medicine 2016;76:163–171.
Scheffler K, Heule R, Báez‐Yánez MG, Kardatzki B, Lohmann G.
The BOLD sensitivity of rapid steady‐state sequences.
Magnetic Resonance in Medicine. 2019; 81:2526–2535.
Nguyen D, Bieri O.
Motion-Insensitive Rapid Configuration Relaxometry.
Magnetic Resonance in Medicine. 2017; 78:518–526.
Heule R, Bause J, Pusterla O, Scheffler K.
Multi‐parametric artificial neural network fitting of phase‐cycled balanced steady‐state free precession data.
Magnetic Resonance in Medicine. 2020; 84:2981–2993.
Birk F, Glang F, Loktyushin A, Birkl C, Ehses P, Scheffler K, Heule R.
High‐resolution neural network‐driven mapping of multiple diffusion metrics leveraging asymmetries in the balanced steady‐state free precession frequency profile.
NMR in Biomedicine, 2021.
Gorgolewski K, Burns CD, Madison C, Clark D, Halchenko YO, Waskom ML, Ghosh SS. Nipype.
A Flexible, Lightweight and Extensible Neuroimaging Data Processing Framework in Python.
Frontiers in Neuroinformatics. 2011; 5.



Go to Editor View