.flow file units - inconsistent in cerebral repository models? (cc/sec?)

[Alain Y]

Hi Everyone!

I've been running many simulations with SimVascular since the beginning of May to use it on my summer research project. I've now created an idealized MCA model with an aneurysm to study.

As I've gone through the documentation over several weeks, I've understood that SimVascular does not have any explicit units and when specifying parameters for simulations, we can input them in either CGS, MGS or SI units.

To remain consistent, I've used the "default" CGS units. However, where I hit a roadblock is with the .flow file. My understanding from the documentation is that if the CGS units are used, the .flow file flow rates are in cc/sec (or mL/sec).

I looked at the cerebral models located on the repository as guidance and they don't seem to be consistent with those units, even though the CGS units were used for the viscosity and density. For example, the "inflow.flow" file for the model "0053_H_CERE_H" has values ranging between 0.147-0.566 cc/sec for the left and right inlets of the vertebral arteries. "0052_H_CERE_H" has used individual flow rates for the left and right vertebral arteries ranging between 0.0455 - 0.172 cc/sec. However, a quick Google search brings up an article that mentions in the Results section of the abstract that the flow rate of the individual inlets can range between 47.4-151.7 mL/min (or 0.79-2.53 cc/sec).

Could anyone provide some clarity on the disparity in values? Understanding this is crucial for inputting the correct units/values for my model parameters or else all calculated data from the simulation will be invalid.

Much appreciated!

PS: Thank you for all the effort put into this software and related projects. Very impressive to say the least!

-Alain

[Alain Y]

Would there be any technical reason to use another unit other than cc/sec for the .flow file?

[David Parker]

Hi Alain,

The data in the flow file matches the assumptions made in the paper referenced for the 0052_H_CERE_H VMR entry.

If your model/mesh does not have a length scale in CGS then you will need to use other units.

Cheers,
Dave

[Alain Y]

Hi Alain,

The data in the flow file matches the assumptions made in the paper referenced for the 0052_H_CERE_H VMR entry.

If your model/mesh does not have a length scale in CGS then you will need to use other units.

Cheers,
Dave

Hi David,

Thanks for getting back to me! Would it be possible to clarify? I feel like there's something I'm not understanding.

The article provides the following measured inflow rates for the patients 1, 2, 3 and 4 (i.e. VMR projects 0051/52/53/54_H_CERE_H respectively). Which is similar to the 2.4 cc/sec I want to simulate.

Table1.png (17.25 KiB) Viewed 243 times

I plotted the data from the four inflow.flow files of these projects below, used to set the flow rates for each of the two inlets (which according to SimVascular documentation should also be in cc/sec):

2023-07-27 11_23_34-Book1.xlsx - Excel.png (44.28 KiB) Viewed 243 times

The flowrate values of the inflow.flow don't range or peak anywhere near the individual, average or sum of the measured VA inlets of each patient mentioned in the table above. As mentioned in the article and when opening these projects in SimVascular, CGS units were used for density and viscocity (1.06 and 0.04). What am I missing?

[Alain Y]

Hi again David,

I found another cerebral model within the newly uploaded repository models (amazing!), that is also a pulsating simulation very similar to my idealized model. Project: 0207_H_CERE_CA

In this case, all the units/decimal points, boundary conditions and flow rates (graph below) matched with what I expected to see:

2023-07-28 14_08_34-Book2.xlsx - Excel.png (10.92 KiB) Viewed 212 times

If you perhaps knew/understood why the other flowrates were so low, that would be great. Just in case it's something I'm missing on my end! If not, no worries.

[David Parker]

Hi Alain,

You were correct in your interpretation of the 0052_H_CERE_H inflow data. I didn't read the referenced paper corrected, thought the data was some sort of average.

I will contact the VMR guys and see if they understand what's going on. Thanks for pointing this out!

Cheers,
Dave

[Luca Pegolotti]

Hi Alain,

Thank you so much for your message and for pointing out that some inflow profiles in the SV projects on the VMR website do not match those reported in the paper. Unfortunately, those models are pretty old, and we couldn't find the actual boundary condition files used in the paper. We will work on updating the models by scaling their inflow profiles to recover the flow rate reported in the reference. I think that we will be able to upload the new models within the next week or so.

Best,
Luca

[Alain Y]

Hi Alain,

You were correct in your interpretation of the 0052_H_CERE_H inflow data. I didn't read the referenced paper corrected, thought the data was some sort of average.

I will contact the VMR guys and see if they understand what's going on. Thanks for pointing this out!

Cheers,
Dave

Hi Alain,

Thank you so much for your message and for pointing out that some inflow profiles in the SV projects on the VMR website do not match those reported in the paper. Unfortunately, those models are pretty old, and we couldn't find the actual boundary condition files used in the paper. We will work on updating the models by scaling their inflow profiles to recover the flow rate reported in the reference. I think that we will be able to upload the new models within the next week or so.

Best,
Luca

Thank you for confirming my doubts David and Luca! As mentioned, I wanted to be sure I wasn't forgetting/neglecting to consider any additional factors in my own simulation. Appreciate it and glad I could help out in the process!

Regards,
-Alain