Masters Thesis

Semi-automatic segmentation of the scalae in the human inner ear, based on parametrization of the cochlear shape

Segmentation of structures in the human cochlea is labor intensive. However, the scalae (Scala Tympani, Scala Media, Scala Vestibuli) show clear structural properties, which could be exploited to generate an appropriate segmentation with little manual labeling of the structures.

A paper by Wimmer et al. (https://link.springer.com/chapter/10.1007/978-3-030-32254-0_1) parametrized human cochleae based on their modiolar axis, rotation, translation and scale, to capture their unique spiral geometry. The task of the masters thesis is to implement the algorithm from Wimmer et al., and use it to generate a complete segmentation of cochlear scalae based on manual segmentation of a few scan slices, by projecting them along the determined spiral parameters.

If you are interested in this topic, write to albert.croner@tum.de, including your current transcript of records and CV.

Masters Thesis

Vibro-tactile stimulation for the desynchronization of periodic brain activity

Vibro-tactile stimulation of the fingertips has been shown to have beneficial effects in patients with Parkinson's disease (Syrkin-Nikolau et al. 2018, Pfeifer et al. 2021). To optimize the stimulation pattern used, it is necessary to study the effects of vibrotactile stimulation on ongoing periodic brain activity.

The task of the work will be to carry out EEG measurements during vibrotactile stimulation in healthy subjects, to investigate the effects of the stimulation on the measured EEG activity and to adapt the stimulation patterns accordingly.

The project is carried out in cooperation with the startup company curetec GmbH, located in Starnberg. You will need to be present at Starnberg 2-3 days a week. Prior knowledge about the physiology of the brain and its activity is desired. Experience in the field of EEG analysis would be desirable and experience in programming with Matlab and Python is required.

If you are interested, write to c.hauptmann@tum.de, including your current transcript of records and CV. Earliest starting date: Mid-June 2024

Research Internship

Nerve + Rosenthal's Canal Segmentation

In cooperation with the Medical University of Innsbruck, we obtained new, high-resolution scans of the human inner ear, scanned in the MuCLS. The scans have a voxel-resolution of 6 µm, and good visibility on the auditory nerve (left figure).

The scans are required for improvements in models of the human cochlea, especially in regard to electrical stimulation by cochlear implants. To make use of the scans in the models, however, first a segmentation of the relevant structures is necessary. We already have semi-automatic segmentations of the auditory nerve (right figure).

The task of the project is the finalization of the current nerve segmentation. Two aspects are especially relevant:
- Fixing small gaps in the nerve segmentation, caused by artifacts in the original scans.
- Segmentation of the Rosenthal's Canal, the structure containing the cell bodies of the neurons in the auditory nerve. The canal is visible as a gap in the current segmentation.

In addition, the morphology of the segmented structures is to be measured and interpreted.

Prior knowledge about the anatomy and functionality of the inner ear and neurons is desired. For this, a visit of the lectures Neuroprosthetics or Systemtheorie der Sinnesorgane is advantageous. The segmentation will be done on Dragonfly or 3DSlicer, experience in the program is advantageous, but not required.

If you are interested, write to albert.croner@tum.de, including your current transcript of records and CV. Earliest starting date: July 2024

Additional projects

For information about additonal topics for project internship, research internship or masters thesis, please refer to info@bai.cit.tum.de, with your CV and transcript of records.