Machine learning of dorsal column nuclei surface potentials evoked by tactile- and proprioceptive-dominated mechanical stimuli (rat data set)

28 Commits

4 Branches

1 Releases

Jason Potas 605c0f64fd Update 'datacite.yml'		vor 4 Jahren
extracted_features	359cc4d659 upload extracted features for six rats	vor 4 Jahren
raw_data	2a033887ab Upload MS07 right side	vor 4 Jahren
LICENSE	7ee29f0378 Initial commit	vor 4 Jahren
README.md	8ada1fcd0d Update 'README.md'	vor 4 Jahren
data_arrangement.pdf	82a041a3f0 updated data_arragement.pdf uploaded	vor 4 Jahren
datacite.yml	605c0f64fd Update 'datacite.yml'	vor 4 Jahren

tactile.proprio-evoked_DCN_potentials

Machine learning of dorsal column nuclei surface potentials evoked by tactile- and proprioceptive-dominated mechanical stimuli (rat data set).

Raw data organisation is schematically described in the pdf document: data_arrangement.pdf

Extracted features from raw data (for machine learning) is described as follows:

Each row is 1 feature. There are 28 features x 7 electrodes = 196
Below the rows of inputs is the 16 output targets in rows 197-212
Each column is a trial
labels for each row are found in the first two columns

Data are contained in .mat files.

For importing to a Python dictionary, see:

https://github.com/wblakecannon/DataCamp/blob/master/05-importing-data-in-python-(part-1)/2-importing-data-from-other-files-types/the-structure-of-mat-in-python.py

https://towardsdatascience.com/how-to-load-matlab-mat-files-in-python-1f200e1287b5

For importing to Julia, see:

https://github.com/JuliaIO/MAT.jl

datacite.yml
Title	Surface potential recordings from rat brainstem dorsal column nuclei in response to tactile and proprioceptive stimuli
Authors	Loutit,Alastair;UNSW Sydney;ORCID:0000-0002-4102-8814 Potas,Jason;UNSW Sydney;ORCID:0000-0002-1973-8211
Description	The raw data are electrical signals recorded from the surface of rat brainstem using a 7-electrode surface array. Rats were urethane anaesthetised and 4 different tactile and proprioceptive stimuli were applied to each of the four limbs while the electrical brainstem signals were simultaneously recorded. Twenty-eight different types of signal features were quantified from these recordings and stored in matrices that can be used as inputs for machine-learning. The features were quantified from 17 different time windows ranging from 20 ms to 1000 ms in length. In total, there is data from six rats (n = 6). For more details and the analysis and discussion of this dataset, see the paper at doi: https://doi.org/10.1101/831164
License	Creative Commons CC0 1.0 Public Domain Dedication (https://creativecommons.org/publicdomain/zero/1.0/)
References	Alastair J Loutit, Jason R Potas (2020). Novel neural signal features permit robust machine-learning of natural tactile- and proprioception-dominated dorsal column nuclei signals. bioRxiv 831164; doi: https://doi.org/10.1101/831164 [doi: https://doi.org/10.1101/831164] (IsDescribedBy)
Funding	Bootes Medical Research Foundation
Keywords	Neuroscience somatosensation electrophysiology brain brainstem neural coding touch proprioception rat in vivo
Resource Type	Dataset

README.md

tactile.proprio-evoked_DCN_potentials