Automated movement analysis of mouse beam walking using neural networks

Sebastian Kloubert 16c17c6575 Upload files to 'Networks/RotatingBeam3.0-Sebastian_Kloubert-2020-07-12/videos' 3 years ago
DLC_results 6e1117cb09 gin commit from AgAswendt-OptiPlex-9020 3 years ago
Manual_Rotating_Beam_Analysis 6e1117cb09 gin commit from AgAswendt-OptiPlex-9020 3 years ago
Networks 16c17c6575 Upload files to 'Networks/RotatingBeam3.0-Sebastian_Kloubert-2020-07-12/videos' 3 years ago
PB_T3_Stroke 6e1117cb09 gin commit from AgAswendt-OptiPlex-9020 3 years ago
R_scripts 6e1117cb09 gin commit from AgAswendt-OptiPlex-9020 3 years ago
Thesis_figures 6e1117cb09 gin commit from AgAswendt-OptiPlex-9020 3 years ago
LICENSE db3cc0a494 Update 'LICENSE' 3 years ago
README.md c14a25914a Update 'README.md' 3 years ago
datacite.yml dd2e6c1af8 Update 'datacite.yml' 3 years ago

README.md

Automated_Rotating_Beam_Analysis

Hello and welcome to the data repository for my Bachelor thesis:

"Automated movement analysis of mouse beam walking using neural networks"

In this repository all data and results obtained during the thesis can be found.

  • Networks: the two DLC artifical neural networks used for the data processing.
  • R-scripts: R-codes used for the post-processing. Ready for application.
  • Thesis_figures: All figures and tables used in the thesis.
  • PB_T3_Stroke: the 20 original stroke-mice videos used for the evaluation in the results.
  • Manual_Rotating_Beam_Analysis: the results of the manual raters and the 20 analysed videos with added frame indices.
  • DLC_results: All pose predictions, pose plots, labeled videos and results of the statistical analysis. For the first results obtained by DLC (Iteration-0), the eight refined networks (25,000-200,000) and the sham treated mice for the further kinematic analysis. For the refined networks the results are divided in the four different video quailities which were applied (25-100 percent resolution)

If you have any further questions, please do not hesitate to contact me: sebastian.kloubert@uni-duesseldorf.de

datacite.yml
Title Automated movement analysis of mouse beam walking using neural networks
Authors Kloubert,Sebastian;University Hospital Cologne
Aswendt,Markus;University Hospital Cologne;ORCID:0000-0003-1423-0934
Wieters,Frederique;University Hospital Cologne
Description This storage contains the whole data and results of the evaluation whether the software package DeepLabCut (DLC) is able to sufficiently label and thus automate the analysis of videos of mice performing the behavior test Rotating Beam. The post-processing was performed via R. The software package DeepLabCut is able to collect all important data for the automated analysis of the behavior test Rotating Beam and has sufficient accuracy to use the collected data to provide results to a scientific study. The comparison to the manual raters showed that the predictions of DeepLabCut are a good generalization not influenced by human bias. This allows to extend the usage to advanced analyses with more complex parameters.
License Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) (https://creativecommons.org/licenses/by-nc/4.0/)
References
Funding
Keywords Neuroscience
DLC
Rotating Beam
Resource Type Dataset