Book/Dissertation / PhD Thesis FZJ-2018-07475

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png
Finite-Difference Time-Domain Simulations Assisting to Reconstruct the Brain's Nerve Fiber Architecture by 3D Polarized Light Imaging



2018
Forschungszentrum Jülich GmbH Zentralbibliothek, Verlag Jülich
ISBN: 978-3-95806-368-6

Jülich : Forschungszentrum Jülich GmbH Zentralbibliothek, Verlag, Schriften des Forschungszentrums Jülich. Reihe Schlüsseltechnologien / Key Technologies 188, IX, 296 S. () = RWTH Aachen, Diss., 2018

Please use a persistent id in citations:  

Abstract: The neuroimaging technique $\textit{Three-dimensional Polarized Light Imaging (3D-PLI)}$ reconstructs the brain’s nerve fiber architecture by transmitting polarized light through histological brain sections and measuring their birefringence. Measurements have shown that the polarization-independent transmitted light intensity ($\textit{transmittance}$) depends on the out-of-plane inclination angle of the nerve fibers. Furthermore, the optical anisotropy that causes the birefringence leads to polarization-dependent attenuation of light ($\textit{diattenuation}$), which might provide additional information about the underlying fiber configuration. In this thesis, analytical considerations, supplementary measurements, and numerical simulations were performed to study the transmittance and diattenuation effects in more detail, and to develop ideas how the effects can assist the nerve fiber reconstruction with 3D-PLI. The propagation of the polarized light wave through the brain tissue was modeled by $\textit{Finite-Difference Time-Domain (FDTD)}$ simulations. Following a bottom-up approach, the simplest possible model was identified that describes the observed transmittance and diattenuation effects. The experimental studies in this work have shown that the transmittance significantly decreases with increasing inclination angle of the fibers (by more than 50 %). The FDTD simulations could model this effect and show that the decrease in transmittance is mainly caused by polarization-independent light scattering in combination with the limited numerical aperture of the imaging system. Moreover, the simulations revealed that the transmittance does not depend on the crossing angle between horizontal fibers. Combining the simulation results with experimental data, it could be demonstrated that the transmittance can be used to distinguish between horizontal crossing and vertical fibers, which is not possible in standard 3D-PLI measurements. To study the diattenuation of brain tissue, a measurement protocol has been developed that allows to measure the diattenuation even with a low signal-to-noise ratio: $\textit{Diattenuation Imaging (DI)}$. The experimental studies in this work revealed that the diattenuation of brain tissue is relatively small (less than 10%) and that it has practically no impact on the measured 3D-PLI signal. More importantly, it was demonstrated that there exist two different types of diattenuation that are specific to certain fiber configurations: in some brain regions, the transmitted light intensity becomes maximal when the light is polarized parallel to the nerve fibers (D$^{+}$), in other brain regions, it becomesminimal (D$^{−}$). The FDTD simulations could successfully model the diattenuation and show that diattenuation of type D$^{−}$ is caused by anisotropic scattering of light which decreases with increasing time after tissue embedding, while diattenuation of type D$^{+}$ can be caused both by anisotropic scattering and by anisotropic absorption (dichroism). In addition, the simulations confirmed that steep fibers only show diattenuation of type D$^{+}$ and that the diattenuation also depends on the tissue composition. This makes Diattenuation Imaging a promising imaging technique that reveals different types of fibrous structures which cannot be distinguished with current imaging techniques.


Note: RWTH Aachen, Diss., 2018

Contributing Institute(s):
  1. Strukturelle und funktionelle Organisation des Gehirns (INM-1)
Research Program(s):
  1. 574 - Theory, modelling and simulation (POF3-574) (POF3-574)

Appears in the scientific report 2018
Database coverage:
Creative Commons Attribution CC BY 4.0 ; OpenAccess
Click to display QR Code for this record

The record appears in these collections:
Institute Collections > INM > INM-1
Document types > Theses > Ph.D. Theses
Document types > Books > Books
Workflow collections > Public records
Publications database
Open Access

 Record created 2018-12-17, last modified 2022-09-30