English
 
User Manual Privacy Policy Disclaimer Contact us
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT

Released

Paper

Combining deep learning and active contours opens the way to robust, automated analysis of brain cytoarchitectonics

MPS-Authors
/persons/resource/persons226758

Thierbach,  Konstantin
Department Neurophysics (Weiskopf), MPI for Human Cognitive and Brain Sciences, Max Planck Society;

/persons/resource/persons226760

Gavriilidis,  Filippos
Department Neurophysics (Weiskopf), MPI for Human Cognitive and Brain Sciences, Max Planck Society;

/persons/resource/persons180061

Kirilina,  Evgeniya
Department Neurophysics (Weiskopf), MPI for Human Cognitive and Brain Sciences, Max Planck Society;

/persons/resource/persons213123

Jäger,  Carsten
Department Neurophysics (Weiskopf), MPI for Human Cognitive and Brain Sciences, Max Planck Society;

/persons/resource/persons19656

Geyer,  Stefan
Department Neurophysics (Weiskopf), MPI for Human Cognitive and Brain Sciences, Max Planck Society;

/persons/resource/persons147461

Weiskopf,  Nikolaus
Department Neurophysics (Weiskopf), MPI for Human Cognitive and Brain Sciences, Max Planck Society;

/persons/resource/persons201756

Scherf,  Nico
Department Neurophysics (Weiskopf), MPI for Human Cognitive and Brain Sciences, Max Planck Society;

Fulltext (public)

Thierbach_2018.pdf
(Preprint), 3MB

Supplementary Material (public)
There is no public supplementary material available
Citation

Thierbach, K., Bazin, P.-L., De Back, W., Gavriilidis, F., Kirilina, E., Jäger, C., et al. (2018). Combining deep learning and active contours opens the way to robust, automated analysis of brain cytoarchitectonics. BioRxiv. doi:10.1101/297689.


Cite as: http://hdl.handle.net/21.11116/0000-0002-130B-D
Abstract
Deep learning has thoroughly changed the field of image analysis yielding impressive results whenever enough annotated data can be gathered. While partial annotation can be very fast, manual segmenta- tion of 3D biological structures is tedious and error-prone. Additionally, high-level shape concepts such as topology or boundary smoothness are hard if not impossible to encode in Feedforward Neural Networks. Here we present a modular strategy for the accurate segmentation of neural cell bodies from light-sheet microscopy combining mixed-scale convolutional neural networks and topology-preserving geometric deformable models. We show that the network can be trained efficiently from simple cell centroid annotations, and that the final segmentation provides accurate cell detection and smooth segmentations that do not introduce further cell splitting or merging.