Keywords: sHybridNeuro, webinar, teaching materials, statistical models, regression analysis, motor unit identification, matlab, rstudio, statistics, surface high density electromyogram (HDEMG), tibialis anterior, dataset

SHARING/ACCESS INFORMATION
This dataset is available under the Creative Commons Public Domain Dedication (CC0) license.

Publicly accessible location of the data: https://dk.um.si/IzpisGradiva.php?lang=eng&id=88845

Video recording of the webinar is available on YouTube: https://www.youtube.com/watch?v=JoKu6kqnEoc

Recommended citation for this dataset:
Holobar, Aleš, Murks, Nina. “Materials for HybridNeuro webinar titled "Validation of results: statistical models and MU identification accuracy"”. System Software Laboratory, Faculty of Electrical Engineering and Computer Science, University of Maribor, Slovenia. Digital library of University of Maribor. https://dk.um.si/IzpisGradiva.php?lang=eng&id=88845

DATA & FILE OVERVIEW
This dataset contains a collection of teaching materials that were used in the HybridNeuro project webinar titled "Validation of results: statistical models and MU identification accuracy". The webinar was presented by Aleš Holobar and covered the complexities of motor unit (MU) identification accuracy, regression analysis and Bayesian models. The primary aim of the webinar was to spark a robust discussion within the scientific community, particularly focusing on the application and implications of linear mixed models and Bayesian regression in the realm of MU identification. The teaching materials include Matlab and R source code for statistical analysis of the included data, as well as three examples of MU identification results in CSV format (from both synthetic and experimental HDEMG signals). The presentation slides in PDF format are also included. The dataset is approximately 9 MB in size.

File list:

• README.txt: this text file containing description of the dataset
• metadata.xml: Dublin Core metadata in XML format
• folder "data":
•      - DATA_ExperimentalEMG.csv: data table with metrics calculated from decomposed experimental HDEMG signals, in CSV format
•      - DATA_SyntheticEMG.csv: data table with metrics calculated from decomposed synthetic HDEMG signals, in CSV format, a shorter example
•      - LARGE_DATA_SyntheticEMG.csv: data table with metrics calculated from decomposed synthetic HDEMG signals, in CSV format, a longer example
• folder "matlab_code":
•      - MATLAB_code_ExperimentalEMG.m: Matlab code for statistical analysis of the experimental data
•      - MATLAB_code_SyntheticEMG.m: Matlab code for statistical analysis of the synthetic data
• folder "R_code":
•      - R_code_ExperimentalEMG.R: R code for statistical analysis of the experimental data
•      - R_code_SyntheticEMG.R: R code for statistical analysis of the synthetic data
• HybridNeuro_Webinar_Holobar_ValidationOfResults_slides.pdf: presentation slides used during the webinar, in PDF format

METHODOLOGICAL INFORMATION
Synthetic EMG decomposition results are based on simulated surface HDEMG signals generated by a multilayer cylindrical volume conductor model [1] for the tibialis anterior muscle. The MU firing pattern was computed by the model proposed in [2] with the parameters adapted to the tibialis anterior muscle. Signals were simulated for 15 subjects performing the 30 % ramp contractions, with 3 repetitions per subject and 15 simulated motor units.

Experimental EMG decomposition results are based on surface HDEMG recordings from 19 healthy volunteers. HDEMG signals were recorded from tibialis anterior muscle during 30 % ramp contractions using a 13x5 electrode grid (GR08MM1305, OT Bioelettronica, Torino, Italy), resulting in 64 data channels sampled at 2048 Hz and 16 bit resolution.

Both recorded as well as simulated signals were then decomposed into sets of motor units and their firing instants by using the DEMUSE Tool software (https://demuse.feri.um.si/). Decomposition results were manually checked and edited by two experts and only accurately identified motor units were kept for final analysis. Several metrics were calculated from decomposed data and stored in CSV files: discharge rate, recruitment threshold, etc. The data in those CSV files serves as the basis for the statistical analysis performed in this webinar, which is included as the Matlab and R source code.

Instrument- or software-specific information needed to interpret the data:
To run the Matlab code the Matlab software is required (https://www.mathworks.com). Alternatively, Matlab code can be run using the GNU Octave software which is an open source alternative to Matlab (https://octave.org/), however the included code was not verified to run correctly in Octave.

For running R code we suggest using the open source RStudio IDE software (https://posit.co/download/rstudio-desktop/).

For generation of synthetic sEMG signals we used the multilayer cylindrical volume conductor model by D. Farina et al. [1].

For decomposition of sEMG signals we used the DEMUSE Tool software, version 6.3, University of Maribor, Slovenia, available at: https://demuse.feri.um.si/

Quality-assurance procedures performed on the data:
The results of decomposition were manually inspected by an expert and any motor units that had pulse-to-noise ratio (PNR) lower than 28 dB [3] were discarded in order to ensure high accuracy of motor unit firing identification.

DATA-SPECIFIC INFORMATION FOR SYNTHETIC EMG RESULTS IN CSV FILES
CSV files contain 6 data colums:

• SubID: integer number. ID label of the subject.
• ContRep: integer number. ID label of contraction repetition.
• MUid: integer number. ID label of the identified motor unit.
• RecThr: double number. Motor unit recruitment threshold, as % of muscle excitation.
• RecDR: double number. Motor unit discharge rate at recruitment (Hz).
• MaxDR: double number. Maximal motor unit discharge rate.

DATA-SPECIFIC INFORMATION FOR EXPERIMENTAL EMG RESULTS IN CSV FILE
CSV files contain 8 data colums:

• SubID: integer number. ID label of the subject.
• ContLev: integer number. Contraction level.
• ContRep: integer number. ID label of contraction repetition.
• MUid: integer number. ID label of the identified motor unit.
• RecThr: double number. Motor unit recruitment threshold, as % of muscle excitation.
• RecDR: double number. Motor unit discharge rate at recruitment (Hz).
• MaxDR: double number. Maximal motor unit discharge rate.
• PNR: double number. Estimated pulse-to-noise ratio in dB.

REFERENCES:
[1] Farina, Dario, Luca Mesin, Simone Martina, and Roberto Merletti. "A surface EMG generation model with multilayer cylindrical description of the volume conductor." IEEE Transactions on Biomedical Engineering 51, no. 3 (2004): 415-426. doi: 10.1109/TBME.2003.820998.

[2] Fuglevand, Andrew J., David A. Winter, and Aftab E. Patla. "Models of recruitment and rate coding organization in motor-unit pools." Journal of neurophysiology 70, no. 6 (1993): 2470-2488.

[3] Holobar, Aleš, Marco Alessandro Minetto, and Dario Farina. "Accurate identification of motor unit discharge patterns from high-density surface EMG and validation with a novel signal-based performance metric." Journal of neural engineering 11, no. 1 (2014): 016008.