[SOLVED] Creating models of Arteries and Veins with sofa

[Garibalde]

How would i create an efficient stable model of arteries and veins in SOFA. Is there a special type of model to use. Tetrahedron models are too expensive. I have tried to use stiff spring linked together…however these have been unstable when constrained.

Any other suggestions or ideas?

Thanks
Vinay

[Hugo]

Hi @vmenon,

To represent blood vessels:

• triangular surfaces are usually used when modeling endovascular catheter navigation, see this paper
• or beams when modeling the vessels as a mechanical structure, see this paper

For the first point, the centerline of vessels can be semi-automatically extracted from medical images, and the triangular mesh can be generated. A plugin will soon be available for licensing or research collaboration on the SOFA marketplace.
Hope this helps,

Hugo

[Garibalde]

I was looking for a simple representation. I have used a simulated chain of spheres connected via JointSpringForceField to produce a representation of a segment or an artery or vein. I used SkinningMapping for the visual node. This seems to work in general however i do have stability issues at times when the end point of the segment are constrained.

So in summary this is my layout of the segment.

<MechanicalObject template=”Rigid” name=”DOFs”  restScale=”1″  position=”Insert here positions”>
<UniformMass template=”Rigid” name=”mass”  mass=”1 1 [1 0 0,0 1 0,0 0 1]” />
<FixedConstraint template=”Rigid” name=”fixOrigin”  indices=”0 10″ />
<UncoupledConstraintCorrection compliance=”0.01″/>
<Node name=”Seg_Test_Vinay.subnode”>
    <MechanicalObject template=”Rigid” name=”attaches”  restScale=”1″  position=”Insert here positions”>
    <RigidRigidMapping template=”MechanicalMapping<Rigid,Rigid>” name=”default6″  repartition=”1 2 2 2 2 2 2 2 2 2 1″ input=”@../DOFs”  output=”@attaches”/>
    <JointSpringForceField template=”Rigid” name=”joints”  spring=”Define spring between nodes”>

    <Node name=”Visu” tags=”Visual”>
        <OgreVisualModel  name=”VisualModel” fileMesh=”mesh/Seg_Cut_Test.obj” />
        <SkinningMapping name=”visual mapping” input=”@../../DOFs” output=”@VisualModel” />
    </Node>

    <Node name=”Surf”>
        <SphereLoader filename=”mesh/Seg_Cut_Test.sph”/>
        <MechanicalObject name=”spheres” translation=”0.0 0.0 0.0″ rotation=”0.0 0.0 0.0″ position=”@[-1].position” />
        <TSphereModel name=”CollisionModel” listRadius=”@[-2].listRadius”  collisionBodyId=”96″/>
        <SkinningMapping name=”sphere mapping” input=”@../../DOFs” output=”@spheres” />
    </Node>

Something like this would work for my use. I need it to be simple, computationally friendly and contactable (collision possible). Do you ahve something that is more efficient or better in SOFA currently that i could use?

Thanks
Vinay

[Hugo]

Did you take a look at the example : examples/Components/forcefield/BeamFEMForceField.scn ?

Hugo

[Wong]

Hello @hugo,
1)I have read the paper and the paper mentions a mapping between a tetrahedron and a 6DoF-beam node. Does such a mapping exist in current Sofa version?

2)I have read the BeamFEMForceField.scn and it runs fine. But when I add the FreeMotionAnimationLoop component into this scene, as shown below, the collision model behaves unnormally.

<Node name="root" dt="0.005">
    <RequiredPlugin pluginName="SofaMiscCollision" />
    <VisualStyle displayFlags="showBehaviorModels showForceFields showCollisionModels" />
    <CollisionPipeline depth="6" verbose="0" draw="0" />
    <BruteForceDetection name="N2" />
    <CollisionResponse name="response" response="FrictionContact" />
    <LocalMinDistance name="proximity" alarmDistance="0.15" contactDistance="0.05" angleCone="0.0" />
    <FreeMotionAnimationLoop/>
    <LCPConstraintSolver tolerance="0.001" maxIt="1000"/>
    <!--<CollisionGroup name="Group" />-->
    <Node name="beam">
        <EulerImplicit rayleighStiffness="0" printLog="false"  rayleighMass="0.1" />
        <BTDLinearSolver template="BTDMatrix6d" printLog="false" verbose="false" />
        <MechanicalObject template="Rigid" name="DOFs" size="8" position="0 0 0 0 0 0 1  1 0 0 0 0 0 1  2 0 0 0 0 0 1  3 0 0 0 0 0 1  4 0 0 0 0 0 1  5 0 0 0 0 0 1  6 0 0 0 0 0 1  7 0 0 0 0 0 1" />
        <Mesh name="lines" lines="0 1 1 2 2 3 3 4 4 5 5 6 6 7" />
        <FixedConstraint name="FixedConstraint" indices="0" />
        <UniformMass mass="1 1 0.01 0 0 0 0.1 0 0 0 0.1" printLog="false" />
        <BeamFEMForceField name="FEM" radius="0.1" youngModulus="20000000" poissonRatio="0.49"/>
        <PrecomputedConstraintCorrection recompute="1"/>
        <Node name="Collision">
            <!--        <CubeTopology nx="115" ny="4" nz="4" xmin="0" xmax="7" ymin="-0.1" ymax="0.1" zmin="-0.1" zmax="0.1" /> -->
            <CubeTopology nx="15" ny="2" nz="2" min="0 -0.1 -0.1" max="7 0.1 0.1" />
            <MechanicalObject />
            <BeamLinearMapping isMechanical="true" />
            <Triangle />
        </Node>
    </Node>
</Node>

3) The blood vessel I want to represent is not an ordinary one. It is an artery with an aneurysm. So what is the best way to simulate such a blood vessel?

[Hugo]

Hi @outtt,

1) I guess the mapping used was the BarycentricMapping, but not being one of the authors, I can’t tell for sure
2) after a short investigation, it appears that the PrecomputedConstraintCorrection is responsible for the instability that you noticed. I am no expert from this code, but it seems the case recompute=”0″, is not really handled. Since you use a UniformMass (constant and diagonal mass matrix), I would rather go for a UncoupledConstraintCorrection
3) it depends on what you expect from your artery model. What is your targeted simulation?

Best

Hugo

[Wong]

Hello @hugo,
Thank you, I have solved the problem.

[Author]

Posts