Projekt 01FMTHH17

Improving quality of spinal cord DTI using inpainting

Ausgangssituation und Zielsetzung

The goal of this multi-disciplinary proposal is to build upon the existing collaboration between TUHH and UKE with the aim to refine inpainting of diffusion tensor imaging (DTI) data such that it reliably and automatically identifies artefactual voxels in clinical spinal cord DTI data, and re- places them with corrected values. An important step in the identification process is an adequate image segmentation mechanism that can improve the detection of artefacts.

Vorgehensweise und Methoden

First, the error of the diffusion tensor-model fit will be used to automatically identify artefactual data points (an approach that has been previously used in outlier-rejection methods). To improve the segmentation of artefactual data points we will use convolutional neural networks and deep learning. Then, the inpainting algorithm will be used to replace the artificial voxels using information from the adjacent voxels, instead of removing these data points, thereby increasing the noise level (as done in outlier-rejection).


The proposed algorithm was implemented and validated using simulation data. On the basis of this project, a new collaborative research project between the UKE and TUHH was created: automated segmentation of 2D and 3D objects using analytical and deep-learning methods, including one publication (Tabarin et al., 2019), one internship project (Najafi), a bachelor (B.Sc: Klisch), a master thesis (M.Sc: Przybyla) and two PhD projects (PhD: Ashtarayeh and Mordhorst). Part of these developments was the consideration of geometric deep learning, a strategy that extends convolutional neural networks to data defined on graphs (Klisch and Przybyla). This step is an additional component for an adequate detection of artefacts in more complex settings. Moreover, two studies, one on spinal cord imaging (David et al., 2019) and one on artefact removal (Papazo- glou et al., 2019), were published during the funding period of this project.


Dr. Siawoosh Mohammadi
Beispiel Projektleitung
Beispiel Institut für X
Beispiel TUHH oder UKE oder X

Prof. Dr. Marko Lindner
Beispiel Projektleitung
Beispiel Institut für X
Beispiel TUHH oder UKE oder X

Dr. Mijail Guillemard
Beispiel Projektleitung
Beispiel Institut für X
Beispiel TUHH oder UKE oder X

Drittmittelprojekte und Drittmittelanträge

  • A BMBF grant (01EW1711A, €150k) was successfully collected in 2019 to ensure the prolongation of the above mentioned projects.


Peer-reviewed publications

  • David G, Mohammadi S*, Martin AR, Cohen-Adad J, Weiskopf N, Thompson A, Freund P (2019) Traumatic and nontraumatic spinal cord injury: pathological insights from neuroimaging. Nature Reviews Neurology:1–14.
  • Papazoglou S, Streubel T, Ashtarayeh M*, Pine KJ, Edwards LJ, Brammerloh M, Kirilina E, Morawski M, Jäger C, Geyer S, Callaghan MF, Weiskopf N, Mohammadi S* (2019) Biophysically motivated efficient estimation of the spatially isotropic component from a single gradient-re- called echo measurement. Magnetic Resonance in Medicine 82:1804–1811.
  • Tabarin T*, Morozova M, Jaeger C, Rush H, Morawski M, Geyer S, Mohammadi S* (2019) Deep learn-ing segmentation (AxonDeepSeg) to generate axonal-property map from ex vivo human optic chiasm using light microscopy. In: Proc Intl Soc Magn Reson Med. 2019;28: #4722.*Authors supported from this grant.

PhD/Master/Bachelor theses supervised as part of this grant proposal

  • Christoph Nicolai (Technomathematik M.Sc. thesis.) Juli 4th 2018. Image Segmentation Methods and an Application to Brain Images Techniques: Mumford-Shah functional, optimization of varia- tional problems.
  • Daniel Klisch (Computer Science, B.Sc. thesis.) October 18th, 2018. Analyzing MRI Data using Geometric Deep Learning.
  • Björn Przybyla (Technomathematik Masterarbeit) WiSe 19/20. Geometric Deep Learning and Applications to Medical Image Analysis.
  • Ruhullah Najafi (Informatik Projektarbeit) WiSe 19/20. UNET and applications to Medical Image Analysis.
  • Mohammad Ashtarayeh (PhD) 2018-2021. Biophysical modelling of the MR signal and validation.
  • Laurin Mordhorst (PhD) 2019-2022. 2D and 3D segmentation using deep-learning.