Question on Simulating Crimping, Deployment, and Cardiac Cycle in TAVR Modeling

[皮卡丘卡皮卡丘]

Dear Prof,

I’ve been carefully studying the excellent PhD dissertation titled:
"Modeling Transcatheter Aortic Valve Replacement in Patient-Specific Anatomies: Fluid-Structure Interaction Models Using the Immersed Finite Element-Difference Method",
and I’m truly impressed by the detailed simulation of the three key stages of TAVR: crimping, deployment, and cardiac cycle.

However, I have several questions regarding how these stages were implemented:

Are all three stages part of the same simulation (i.e., within the same program and timeline), or are they modeled as separate simulations? From the dissertation and supplemental video, it appears that crimping may be performed separately (without the vessel), while deployment and the cardiac cycle are simulated together. Is this understanding correct?

If these are indeed separate simulations, how is the crimped stent configuration transferred to the deployment simulation? That is, how is the deformed geometry (and possibly stress state) of the stent after crimping carried over?

The text mentions that the stent self-expands upon sheath removal due to the restoring forces in the hyperelastic model (i.e., the stress-free configuration is the fully expanded shape). So I wonder: Is the deployment simulation initialized with the crimped geometry but still associated with the expanded configuration as its stress-free state? If so, how is this achieved in practice in IBAMR?

If possible, could you also kindly share any suggestions or best practices that might help me implement this workflow more effectively? Thank you so much for your time and help!

Best regards,

Peng

[皮卡丘卡皮卡丘]

Hi all,

Just following up on my earlier question regarding the TAVR simulation stages (crimping, deployment, cardiac cycle) in IBAMR.

After further reading and studying the dissertation, I feel that the process is likely divided into two stages: 1.crimping, and 2. deployment combined with the cardiac cycle.

I wonder if this understanding is correct. I also think a reasonable approach might be to carry over the deformed mesh from the crimping simulation—specifically the deformed configuration (dX_1, dX_2, dX_3)—when initializing the deployment simulation. But I’m currently stuck on how to properly identify or load this deformation when starting the second simulation.

Any suggestions or guidance on how to implement this step would be greatly appreciated.

Best,
Peng

[Boyce Griffith]

Sorry for the slow response. Some replies follow below:

On Jul 15, 2025, at 1:28 AM, '皮卡丘卡皮卡丘' via IBAMR Users <ibamr...@googlegroups.com> wrote:

Hi all,

Just following up on my earlier question regarding the TAVR simulation stages (crimping, deployment, cardiac cycle) in IBAMR.

After further reading and studying the dissertation, I feel that the process is likely divided into two stages: 1.crimping, and 2. deployment combined with the cardiac cycle.

It is possible to disable structures in IBAMR, so that they do not interact with the fluid (and hence do not move).

What we did was to load all of the structures but to disable the valve apparatus during crimping, and then to disable the device model to open up the native aortic valve, and then enable the TAVR model to deploy it.

I wonder if this understanding is correct. I also think a reasonable approach might be to carry over the deformed mesh from the crimping simulation—specifically the deformed configuration (dX_1, dX_2, dX_3)—when initializing the deployment simulation. But I’m currently stuck on how to properly identify or load this deformation when starting the second simulation.

This should be possible — you can save the state for an FE model in an output file, like an ExodusII file — but I think it will require some coding to load it back in. This could be a nice feature to add to the library in case you are able to do it.

— Boyce

Any suggestions or guidance on how to implement this step would be greatly appreciated.

Best,
Peng

在2025年7月9日星期三 UTC+8 16:16:50<皮卡丘卡皮卡丘> 写道：

Dear Prof,

I’ve been carefully studying the excellent PhD dissertation titled:
"Modeling Transcatheter Aortic Valve Replacement in Patient-Specific Anatomies: Fluid-Structure Interaction Models Using the Immersed Finite Element-Difference Method",
and I’m truly impressed by the detailed simulation of the three key stages of TAVR: crimping, deployment, and cardiac cycle.

However, I have several questions regarding how these stages were implemented:

Are all three stages part of the same simulation (i.e., within the same program and timeline), or are they modeled as separate simulations? From the dissertation and supplemental video, it appears that crimping may be performed separately (without the vessel), while deployment and the cardiac cycle are simulated together. Is this understanding correct?

If these are indeed separate simulations, how is the crimped stent configuration transferred to the deployment simulation? That is, how is the deformed geometry (and possibly stress state) of the stent after crimping carried over?

The text mentions that the stent self-expands upon sheath removal due to the restoring forces in the hyperelastic model (i.e., the stress-free configuration is the fully expanded shape). So I wonder: Is the deployment simulation initialized with the crimped geometry but still associated with the expanded configuration as its stress-free state? If so, how is this achieved in practice in IBAMR?

If possible, could you also kindly share any suggestions or best practices that might help me implement this workflow more effectively? Thank you so much for your time and help!

Best regards,

Peng

-- 
You received this message because you are subscribed to the Google Groups "IBAMR Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ibamr-users...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ibamr-users/132d5bf8-3de9-4640-8403-f8e776f27aa9n%40googlegroups.com.