One_dimensional central line extraction

[Sajeda Akter]

Hello Developers,
Hope you are doing well.

I am very much interested about the 1D coronary simulation.
I would like to know about center line extraction of a coronary artery model. I have used some models for 3D blood flow simulations but when I use same models for center line extraction then I get some correct and some wrong center lines (Please see attached images).

Could you please anyone suggest me how can I get rid from the problems.

Thank you in advance ...

[David Parker]

Hi Sajeda,

The centerlines extraction computation is very sensitive to the geometry of the model. Finer meshes typically work better. The models must also define a closed surface with no holes and well-defined geometry: no intersecting or zero area polygons. If you upload your model someplace I can download I'll have a look.

Cheers,
Dave

[Sajeda Akter]

Thank you Dave for your quick reply.

I will try by using finer mesh.
I have another query that I see there are several .dat files (e.g. _pressure.dat, _flow.dat). What sorts of information there and how can I visualize these files in paraview.

Thank you in advance

Cheers
Sajeda

[David Parker]

Hi Sajeda,

The bct.dat file is created by the svpre command run from SV when you hit the Create Data Files for Simulation button, it contains time-dependent velocity vectors at the inflow face. The other .dat files contain data and simulation information. Have a look here http://simvascular.github.io/docsFlowSolver.html for an explanation.

You can use Paraview to visualize the time-dependent velocity vectors at the inflow face using the bct.vtp file.

Cheers,
Dave

[Sajeda Akter]

Thank you very much Dave.

Could you please let me know how can I visualize the Group0_Seg0_pressure.dat (e.g. size 47569 KB) file or flow file in Paraview or else in 1D simulation results.

Thank you

Cheers,
Sajeda

[David Parker]

Hi Sajeda,

The best way to visualize 1D simulation results is to convert the results to CSV format (See the 1D Simulation Convert Results panel). You can then use the CSV files to plot retsults.

Cheers,
Dave

[Sajeda Akter]

Hi Dave,
Hope you doing well.
I have a query about the 1D simulation. I have different 3D coronary models with different areas of stenosis. At first, I extracted centerlines from 3D models then I set different materials for 1D simulations. I have observed that all CSV pressure or flow or WSS file produce same results although there were different sizes of stenosis in 3D models. Are there any set up problems or else??

Note: I need different pressure results for different sizes of stenosis in 1D simulation like 3D.

Thanks in advance

Cheers
Sajeda

[Martin Pfaller]

The current default behavior of the 1D modeling pipeline is to place only two nodes within each vessel branch: one at the beginning and one at the end. That means that the cross-sectional area in your vessel is linearly interpolated within those two values. However, as you noticed, that ignores any changes in cross-sectional area along the branch, such as stenoses or aneurysms.

There is a (beta) option to increase the number of 1D nodes per branch. Unfortunately, this option is not accessible through the SimVascular GUI. Instead, you need to call the 1D pipeline from Python. Please find some instructions here: viewtopicPhpbb.php?f=188&t=12659&p=36177&start=0&view= The options you need are:

- seg_min_num: minimum number of segments per branch
- seg_size: segment size
- seg_size_adaptive: if set to True, places 1D nodes automatically in regions with high changes in cross-sectional area

Please note that despite the fine discretization of the vessel, the 1D solver still assumes a Poiseuille flow. This assumption is not valid in regions with high changes in cross-sectional areas and the 1D solver will underestimate the pressure loss. Furthermore, the strong gradients in the cross-sectional area can make the solver unstable. We're working on that to include a loss-model for stenoses in the future.

Let me know if you run into any problems!

[Sajeda Akter]

Thank you very much Martin.
I am very happy to receive your reply.
I will try follow your instructions to measure pressure loss in the downstream of the stenosis region. However, I would be happy if you write some more instructions or python codes in this regards.

Thank you in advance

Cheers,
Sajeda