TU Berlin

Neuronale InformationsverarbeitungKonfokale Mikroskipie: Segmentierung, "Tracing" und Analyse von 3D Bildern

Neuronale Informationsverarbeitung


zur Navigation

Confocal Microscopy: Semi-Automatic Segmentation, Tracing and Analysis of 3D Images


In this project we developed algorithms for the computer assisted segmentation and 3D-reconstruction of neurons from confocal microscope image stacks. We investigated methods for the correction of scaling artifacts due to refractive index mismatch and tissue shrinking. We also developed blind deconvolution techniques in order to comensate for the strongly anisotropic point-spread functions measured in the stained preparations. Deconvolution techniques were validated in preparations of optic neurpils where the resolution of the confocal microscope scans could be sufficiently enhanced in order to study colocalization between synaptic vesicle markers near the resolution limit of light. We evaluated techniques based on the wavelet transform for increasing the signal-to-noise ratio of the confocal images, and we developed semi-automatic algorithms for segmentation, tracing and reconstruction of connected tubular structures. 3D-reconstruction techniques were evaluated on 3D scans of neurons from Maduca Sexta.  

Acknowledgement: Research was funded by BMBF, DFG, and the Technische Universität Berlin.

Ausgewählte Publikationen:

New Methods for the Computer-Assisted 3D Reconstruction of Neurons from Confocal Image Stacks
Zitatschlüssel Schmitt2004
Autor Schmitt, S. and Evers, J.-F. and Duch, C. and Scholz, M. and Obermayer, K.
Seiten 1283 – 1298
Jahr 2004
DOI doi:10.1016/j.neuroimage.2004.06.047
Journal Neuroimage
Jahrgang 23
Nummer 4
Verlag Elsevier
Zusammenfassung Exact geometrical reconstructions of neuronal architecture are indispensable for the investigation of neuronal function. Neuronal shape is important for the wiring of networks, and dendritic architecture strongly affects neuronal integration and firing properties as demonstrated by modeling approaches. Confocal microscopy allows to scan neurons with submicron resolution. However, it is still a tedious task to reconstruct complex dendritic trees with fine structures just above voxel resolution. We present a framework assisting the reconstruction. User time investment is strongly reduced by automatic methods which fit a skeleton and a surface to the data, while the user can interact, and thus, keeps full control to ensure a high quality reconstruction. The reconstruction process comprises a successive gain of metric parameters. First a structural description of the neuron is built, including the topology and the exact dendritic lengths and diameters. We use generalized cylinders with circular cross-sections. The user provides a rough initialization by marking the branching points. The axes and radii are fitted to the data by minimizing an energy-functional which is regularized by a smoothness constraint. The investigation of proximity to other structures throughout dendritic trees requires a precise surface reconstruction. In order to achieve accuracy of 0.1 micron and below, we additionally implemented a segmentation algorithm based on geodesic active contours which allows for arbitrary cross-sections and uses locally adapted thresholds. In summary, this new reconstruction tool saves time and increases quality as compared to other methods which have previously been applied to real neurons.
Typ der Publikation Selected:main selected:3dimages selected:publications
Link zur Originalpublikation Download Bibtex Eintrag

Nach oben



Schnellnavigation zur Seite über Nummerneingabe