If you’re looking for open positions in our lab, please follow this link.
- 1 Bachelor/Master Projects
- 1.1 Multilens array for multifocus imaging
- 1.2 Blind SIM processing of experimental data
- 1.3 New sample embedding media:
- 1.4 Clarity protocol with non-toxic substances
- 1.5 New embedding media for Raman microscopy for medical diagnostics
- 1.6 CARS/SARS theory
- 1.7 Inverse modelling for Holography
- 1.8 Blind PSF deconvolution of light-sheet data
- 1.9 Deconvolution framework in the programming language Julia
- 1.10 High speed Raman-microscopy
Multilens array for multifocus imaging
(Master/Physics, supervising: Sapna Shukla, Aurélie Jost)
This project aims at developing a multifocus imaging system (see also the MF-SIM). A rough concept for the realisation of such a new system exists. Part of the project would be an optical simulation using Zemax. Another part would be to build such a system and characterize its performance.
Blind SIM processing of experimental data
(Bachelor/Physics, supervising: Aurélie Jost)
An algorithm for processing structured illumination data was developed in the group. In collaboration with other groups outside Jena, this “blind-SIM” algorithm should now be applied to several biological datasets of interest. The student would improve the user-interface (by Matlab programming) and understand and apply the blind-SIM reconstruction to biological datasets of interest. Previous knowledge about Matlab is required. A master student could also look at extending the existing code to treating 3D structured illumination stacks, which involves combined deconvolution with multiple point spread functions.
New sample embedding media:
(Bachelor or Master/Chemistry, supervising: Martin Reifarth)
Martin Reifarth synthesized a novel medium with an exceptionally high refractive index. Unfortunately this turned out to not be water soluble. Such a water-soluble medium of high-n and low osmolarity would be a fantastic embedding medium for all types of light-microscopy. The aim of this project is to synthesize such a medium.
Clarity protocol with non-toxic substances
(Bachelor or Master/Chemistry, supervising: Dr. Ulrich Leischner)
The aim is to combine the existing clarity sample embedding procedure with new substances.
New embedding media for Raman microscopy for medical diagnostics
(Master Chemistry, supervising: Dr. Ulrich Leischner, Walter Müller)
Some ideas will be tried using different substances that are promising to yield a low Raman intensity. This should be tried out and residual Raman signal will be hopefully eliminated with the help of image processing.
(Master Physics, supervising: Prof. Rainer Heintzmann)
Coherent antistokes Raman and stimulated Raman scattering is an effect that is often used for imaging. However, no unified theory yet exists. We have some ideas about how such a theory could be made. The candidate would get familiar with the matter and try to expand on the ideas with the aim to develop a unified theory about CARS imaging that is easy to describe in pictures.
Inverse modelling for Holography
(Master Physics, supervising : Prof. Rainer Heintzmann, Marie Walde)
Inverse modelling with the help of an existing Matlab toolbox will be applied to simulated and measured data from multi-wavelength holography/holoscopy. This can allow to image large volumes relatively quickly, which can have a medical impact in fields like ophthalmology.
Optical design of light-sheet illumination optics for a range of imaging conditions
(Bachelor or Master Physics, supervising: Dr. Ulrich Leischner)
The aim is to (individually) optimize the optical performance of the light-sheet microscopy illumination system present in the lab for new types of samples, new illumination wavelengths, different index of refraction and larger field of view.
Blind PSF deconvolution of light-sheet data
(Master Physics, supervising: Prof. Rainer Heintzmann)
A deconvolution framework has been programmed in the Heintzmann lab. The most recent addition is the ability to deconvolve data with an unknown point spread function, by estimating its unknown complex valued amplitude in the aperture plane. The aim here is to expand this scheme to the deconvolution of light-sheet data, where the point spread function is often distorted to an unknown degree. In a second step, one can also try to incorporate ideas about spatial variations in the point spread function.
Deconvolution framework in the programming language Julia
(Master Physics/(Computer Science/Informatics), supervising: Dr. Martin Kielhorn, Prof. Rainer Heintzmann)
In recent year a new (functional) programming language has evolved named Julia. It is very fast, versatile (similar to Matlab), easy to use and free. Therefore the aim is to program the ideas contained in an existing deconvolution framework in Julia and optimize and test it. For further speed improvements, this can also be combined with the ability of exploiting the Cuda framework with the help of CudaMat. Mutable arrays, which are not possible in Matlab can also be exploited to this aim
High speed Raman-microscopy
(Master Physics, supervising: Walter Müller)
Raman microscopy is a chemical contrast imagingmethod without complicated staining. Currently, Raman microscopy is limited by the speed of the LASER spot scanning over a certain field of view. For biological samples, the measurement of one spot takes over 100ms and hence, even for a small image of 100×100 pixels, one would need apx. 20 minutes. The aim of this thesis would be to speed this method up to a few seconds.
If you are enrolled in the FSU and are interested in any of the above projects, or have your own research ideas, please contact the appropriate supervisor and Rainer Heintzmann for more details and discussion.