/bioviz

Biorbd visualization toolkit

Primary LanguagePythonMIT LicenseMIT

logo

bioviz is the Python visualizer of the biorbd biomechanical library (https://github.com/pyomeca/biorbd). This library provides filters and analyses for biomechanical data in C++, Python and MATLAB languages.

However, there is no built-in visualization system, therefore this Python module! It allow for a beautiful, yet simplistic visualization of the model in static or dynamic poses.

So, without further ado, let's remove biorbd from the darkness!

License

How to install

There are two main ways to install bioviz on your computer: installing the binaries from Anaconda (easiest) or installing from the sources (requires to download the sources)

Anaconda (For Windows, Linux and Mac)

The easiest way to install bioviz is to download the binaries from Anaconda (https://anaconda.org/) repositories. The project is hosted on the conda-forge channel (https://anaconda.org/conda-forge/bioviz).

After having installed properly an anaconda client [my suggestion would be Miniconda (https://conda.io/miniconda.html)] and loaded the desired environment to install biorbd in, just type the following command:

conda install -c conda-forge bioviz

Please note that this will install all the dependencies of bioviz including biorbd. If you are a developer of biorbd, this is therefore not the prefered method of installation since the biorbd binaries from conda-forge may interfere with those you create.

Installing from the sources

To install from the sources, you simply have to download the source, navigate to the root folder and (assuming your conda environment is loaded if neede) type the following command

python setup.py install

Please note that this method will not install the dependencies for you, therefore you will have to install them manually, including biorbd. Moreover, the setuptools dependencies must be installed prior to the installation in order for it to work.

Dependencies

bioviz relies on several libraries. The most obvious one is biorbd, but pyomeca is also requires and some others. The first hand dependencies (meaning that some dependencies may require other libraries themselves) are: pandas (https://pandas.pydata.org/), numpy (https://numpy.org/), scipy (https://scipy.org/), matplotlib (https://matplotlib.org/), vtk (https://vtk.org/), PyQt (https://www.riverbankcomputing.com/software/pyqt), biorbd (https://github.com/pyomeca/biorbd), pyomeca (https://github.com/pyomeca/pyomeca) and ezc3d (https://github.com/pyomeca/ezc3d). All these can manually be install using (assuming the anaconda environment is loaded if needed) pip3 command or the Anaconda's following command.

conda install pandas numpy scipy matplotlib vtk pyqt biorbd pyomeca ezc3d -cconda-forge

How to use

bioviz is being as minimalistic as possible in order to make it easy to use. Basically, one has to import the library, load the model and you are good to go. The following code shows how to launch the software without any other options.

import bioviz
bioviz.Viz('path/to/model.bioMod').exec()

Please note that the exec() method won't return the hand until the window is closed. If for some reasons, one wants to keep control of the code, they can manually update the GUI using the following snippet. The while loop will then exit when the window is closed.

import bioviz
biorbd_viz = bioviz.Viz('path/to/model.bioMod')
while biorbd_viz.vtk_window.is_active:
    # Do some stuff...
    biorbd_viz.refresh_window()

Interacting with the model

The interaction with the model is fairly easy. Using the sliders on the left hand side of the main window, you can change the position of the model, according to the degrees-of-freedom the model has. Pressing the reset button replaces the model at its default attitude.

Options at loading

If you are interacting with the model using biorbd, you may want to link that model you are interacting with and the one used by bioviz. To do so, you can actually pass the model using the loaded_model parameter of the constructor, like so.

import biorbd
import bioviz
model = biorbd.Model('path/to/model.bioMod')
# Do some stuff with you model...
bioviz.Viz(loaded_model=model).exec()

The element that are shown on the model can be turned off at the loading of the window. The elements are: the mesh of the segments, the global/segments center of mass, the global/segments reference frame, the skin markers, the muscles, the analyses panel. The following code turns everything off, which is a bit useless since nothing will be shown.

import bioviz
bioviz.Viz('path/to/model.bioMod', 
    show_meshes=False,
    show_global_center_of_mass=False, show_segments_center_of_mass=False,
    show_global_ref_frame=False, show_local_ref_frame=False, 
    show_markers=False, 
    show_muscles=False, 
    show_analyses_panel=False
).exec()

Loading a movement

From the GUI

For loading a movement that the model will perform, one has simply to click on the Load movement button and load the previously save movement file. The file must be a numpy array of $n_{DoF} \times n_{Frames}$ matrix.

From the command line

A second method of attaching a movement to the model is to manually attach it to the GUI. To do so, one simply has to call the load_movement function then either call the exec function, or manually update the GUI. The following code shows these two different methods (depending if manually_animate is set to True or False. Please note that the exec() way is preferred.

import numpy as np
import bioviz
manually_animate = False

# Load the model
biorbd_viz = bioviz.Viz('path/to/model.bioMod')

# Create a movement 
n_frames = 200
all_q = np.zeros((biorbd_viz.nQ, n_frames))
all_q[4, :] = np.linspace(0, np.pi / 2, n_frames)

# Animate the model
if manually_animate:
    i = 0
    while biorbd_viz.vtk_window.is_active:
        biorbd_viz.set_q(all_q[i, :])
        i = (i+1) % n_frames
else:
    biorbd_viz.load_movement(all_q)
    biorbd_viz.exec()

Loading external forces

From the command line

import numpy as np
import bioviz

# Load the model
biorbd_viz = bioviz.Viz('path/to/model.bioMod')

# Create a movement
n_frames = 200
all_q = np.zeros((biorbd_viz.nQ, n_frames))
all_q[4, :] = np.linspace(0, np.pi / 2, n_frames)

f_ext = np.zeros((1,6,n_frames))
# fill the origin of the external force
f_ext[0,:3,:] = np.linspace(np.zeros(3), np.ones(3) * 0.1, n_frames).T
# fill the location of the tip of the arrow
f_ext[0,3:,:] = np.linspace(np.ones(3) * 0.2, np.ones(3) * 0.25, n_frames).T

# Animate the model
biorbd_viz.load_movement(all_q)
biorbd_viz.load_experimental_forces(f_ext, segments=None, normalization_ratio=0.2)
biorbd_viz.exec()

Analyses panel

Slowly but surely, it is plan that the visualizer will become a proper GUI of biorbd, allowing for instance to interactively modify a bioMod file or to perform biomechanical analyses. For now, there is one muscle analyses panel which is available.

Muscle analyses panel

By clicking on Muscles in the Analyses pane, the main window should expand showing a simple muscle analyses panel. In this panel you can select the muscle to plot and the DoF (degree-of-freedom) to run the analyses on. By default, the analyses are perform from $-\pi$ to $\pi$. The vertical bar shows the current position of the model. Changing the position's slider will automatically update the plots of the analyses.

The muscle length plot shows the length of the muscle over the range of the select DoF assuming every other DoFs are constant.

The moment arm plot shows the moment arm about the selected DoF over the range of the selected DoF assuming every other DoFs are constant.

The passive forces plot shows the forces of the passive element of the muscle (assuming there are passive elements) over the range of the selected DoF assuming every other DoFs are constant.

Finally, the active forces plot shows the forces of the active element of the muscle (assuming there as active elements) over the range of the selected DoF assuming every other DoFs are constant. Just next to the plot, there is a slider that activate the muscle (from 0% to 100% activation).

How to contribute

You are very welcome to contribute to the project! There are to main ways to contribute.

The first way is to actually code new features for bioviz such as new analyses modules. The easiest way to do so is to fork the project, make the modifications and then open a pull request to the main project.

The second way is to open issues to report bugs or to ask for new features. I am trying to be as reactive as possible, so don't hesitate to do so!

Cite

If you use biorbd (or its GUI bioviz), we would be grateful if you could cite it as follows:

@misc{Michaud2018biorbdViz,
    author = {Michaud, Benjamin and Begon, Mickael},
    title = {bioviz: A vizualization python toolbox for biorbd},
    howpublished={Web page},
    url = {https://github.com/pyomeca/bioviz},
    year = {2018}
}