Synergy musculoskeletal model generated torques

Citation Author(s):
Federica
Damonte
University of Twente
Massimo
Sartori
University of Twente
Submitted by:
Federica Damonte
Last updated:
Thu, 01/09/2025 - 08:50
DOI:
10.21227/s9tw-g074
Data Format:
*.txt;*.bin;*.xml;*.sto
License:
Creative Commons Attribution
17 Views
Categories:
Biomedical and Health Sciences
Keywords:
muscle synergies; musculoskeletal modeling; subject-specific; walking
0
0 ratings - Please login to submit your rating.
ACCESS DATASET CITE SHARE/EMBED

Abstract 

We developed a control framework for an ankle-powered prosthesis that is based on a neuromusculoskeletal model driven by muscle synergies, aimed at mimicking human motor control strategies during cyclic movements.Refer to recent publication for more details on the research goals and methodology [1]. The personalized muscle synergy model computes muscle excitations at each timestamp of every trial using sensory information: gait phase was calculated using real-time ground reaction forces information, speed information was measured from the instrumented treadmill.In this study, both the gait speed and the instantaneous phase information were employed to modulate respectively the weights and the muscle primitives. The gastrocnemius medialis activation signal was used to drive gastrocnemius medialis, gastrocnemius lateralis, and, soleus muscles because of the constraints imposed by the calibration procedure . The joint kinematics, derived from the internal encoder of the prosthesis, is required to compute muscle-tendon kinematics, hence the musculotendon length (Lmt) and moment arm (MA) were calculated using a set of MTU-specific multidimensional B-splines using the personalized musculoskeletal model [2]. For this analysis, we assumed the knee joint was fully extended. Further, we utilized the muscle activation signals and MTU lengths (Lmt) to compute the muscle-tendon forces . The joint torque at the ankle was the product of MTU forces and moment arms (MA); those values were first translated into current which was input to the closed-loop lower control to drive the prosthesis.

The current dataset contains input and output to the subject-specific neuromusculoskeletal model of each subject tested in three different walking speed conditions.

References : 

[1] Damonte, F., Durandau, G., Gonzalez-Vargas, J., Van Der Kooij, H. and Sartori, M., 2023, September. Synergy-driven musculoskeletal modeling to estimate muscle excitations and joint moments at different walking speeds in individuals with transtibial amputation. In 2023 International Conference on Rehabilitation Robotics (ICORR) (pp. 1-6). IEEE.

[2] Sartori, M., Reggiani, M., van den Bogert, A.J. and Lloyd, D.G., 2012. Estimation of musculotendon kinematics in large musculoskeletal models using multidimensional B-splines. Journal of biomechanics, 45(3), pp.595-601.

Instructions: 

 

DATA & FILE OVERVIEW

File list:

Subject01\Calibration_results: contains Spline coefficients(.bin) , subjectCalibrated_EMG.xml (calibrated muscle model), synergy_optimization_output.xml (optimized synergy coefficients)

Subject01\Day2: contains four folders, one for each experimental trial. Each folder contains CEINMS output files. The input information used to generate the output were recorded in custom.sto (gait phases and speed) and ik.sto (ankle_angle) 

Subject02\Calibration_results: contains Spline coefficients(.bin) , S2.xml(calibrated muscle model), synergy_optimization_output.xml(optimized synergy coefficients)

Subject02\Day2: contains three folders, one for each experimental trial. Each folder contains CEINMS output files. The input information used to generate the output were recorded in custom.sto (gait phases and speed) and ik.sto (ankle_angle) 

If data was derived from another source, list source: I am not including the data used for calibration of the model and optimization of the synergy parameters. Results of the calibration are included in the specific folder for each subject

Software- or Instrument-specific information needed to interpret the data, including software and hardware version numbers: 

During the experiment, a workstation (Windows 10 Pro) ran a communication and processing master (TwinCAT, Beckhoff Automation GmbH \& Co. KG, Verl, Germany) as well as the mid-level controller using the open-source software in C++, Calibrated EMG-Informed NMS Modelling Toolbox (CEINMS-RT).

The software running on Twincat was able to send sensory information in real-time to CEINMS-RT to compute ankle torque output.

For CEINMS-RT follow installation instructions and requirements here: https://ceinms-docs.readthedocs.io/en/latest/). 

In order to reproduce the data contained in the folder : -install CEINMS-RT

                                                         -install SynergiesModel_Plugin

                                                         -create a separate folder for each subject following guidelines

                                                         -include spline coefficients for each subject in the specific folder

                                                         -create an execution.xml file to read gait_phase,speed and kinematic from file

                                                         -run CEINMS-RT for each trial  

Required software :  Visual Studio 2017 (min.)

 

Funding Agency: 
Marie Skłodowska-Curie Actions (MSCA) Innovative Training Networks (ITN) - SimBionics (Neuromechanical Simulation and Sensory Feed- back for the Control of Bionic Legs)
Grant Number: 
H2020-MSCA-ITN-2019 - 860850

Dataset Files

LOGIN TO ACCESS DATASET FILES

 

QUESTIONS?