Investigators
Kenneth S. Campbell and Jonathan F. Wenk
Contact info (email)
k.s.campbell@uky.edu
1. Define context(s)
aid in clinical decision making
identify/explore new therapies
Current Conformance Level / Target Conformance Level
Adequate
Primary goal of the model/tool/database
Our project goal is to develop a predictive multiscale model of the heart that will improve understanding of familial cardiomyopathies and that can be used to help screen potential new therapies for cardiac disease. In the future, scientists and clinicians will be able to use the software to help optimize personalized treatment plans for patients who have different types cardiovascular disease.
Biological domain of the model
Cardiac
Structure(s) of interest in the model
sarcomeric protiens, myofiber/collagen, myocardium
Spatial scales included in the model
nanometer thru millimeter
Time scales included in the model
millisecond thru months
2. Data for building and validating the model
| Data for building the model | Published? | Private? | How is credibility checked? | Current Conformance Level / Target Conformance Level |
|---|---|---|---|---|
| in vitro (primary cells cell, lines, etc.) | We are isolating cells from tissue taken from our own animals in order to quantify calcium handling, force, and molecular assays. | We will be able to compare baseline values to the literature. | Adequate | |
| ex vivo (excised tissues) | Yes. We are in the process of harvesting tissue from small animals, which have already been scanned. | We will be able to compare baseline values to the literature. | Adequate | |
| in vivo pre-clinical (lower-level organism or small animal) | Yes. We have imaged small animals and published several papers related to myocardial strain. | Yes. We are still performing imaging on small animals and analyzing the data. | We have compared our results to the literature and found that we are in good agreement. | Adequate |
| in vivo pre-clinical (large animal) | ||||
| Human subjects/clinical | ||||
| Other: ________________________ |
| Data for validating the model | Published? | Private? | How is credibility checked? | Current Conformance Level / Target Conformance Level |
|---|---|---|---|---|
| in vitro (primary cells cell, lines, etc.) | We are isolating cells from tissue taken from our own animals in order to quantify calcium handling, force, and molecular assays. | We will be able to compare baseline values to the literature. | Adequate | |
| ex vivo (excised tissues) | Yes. We are in the process of harvesting tissue from small animals, which have already been scanned. | We will be able to compare baseline values to the literature. | Adequate | |
| in vivo pre-clinical (lower-level organism or small animal) | Yes. We have imaged small animals and published several papers related to myocardial strain. | Yes. We are still performing imaging on small animals and analyzing the data. | We have compared our results to the literature and found that we are in good agreement. | Adequate |
| in vivo pre-clinical (large animal) | ||||
| Human subjects/clinical | ||||
| Other: ________________________ |
3. Validate within context(s)
| Who does it? | When does it happen? | How is it done? | Current Conformance Level / Target Conformance Level | |
|---|---|---|---|---|
| Verification | PhD students and postdocs in the labs of all investigators. All work is reviewed by the senior investigators. | Once a new module is implemented into the main code. | Numerical solutions are verified and benchmarked at each stage of the development process against solutions from simplified systems which include ‘single element’ finite element (FE) models and analytical models (with closed form solutions). The modular structure of the code will provide additional flexibility in terms of debugging and solution verification. | Adequate |
| Validation | PhD students and postdocs in the labs of all investigators. All work is reviewed by the senior investigators. | Once a new module is implemented into the main code. | The multiscale model spans from molecular to organ-level structural scales and encompasses timescales ranging from milliseconds to months. The project has produced experimental data at each of these levels, which directly relate to the model parameters. The simulations can therefore be validated at each point by comparing computed predictions to analogous measurements. For example, at the cell level, unloaded sarcomere shortening profiles and Ca2+ transients generated by the MyoSim code are calibrated against experiments from isolated cells to determine parameters such as thin filament cooperativity. At the organ level, P-V loops and myocardial strain patterns generated by ventricular FE modelsare calibrated with experimental data from MRI to determine parameters defining the stress-strain relationship, as well as circulatory model parameters. | Adequate |
| Uncertainty quantification | PhD students and postdocs in the labs of all investigators. All work is reviewed by the senior investigators. | Once a new module is implemented into the main code. | In the context of MRI data, we perturb the strain fields in order to quanify how the predicted material parameters change. Since there is inherently some noise in the image data, this helps quanitfy its effect. | Adequate |
| Sensitivity analysis | PhD students and postdocs in the labs of all investigators. All work is reviewed by the senior investigators. | Once a new module is implemented into the main code. | We are using techniques such as Sobol indices to quantify the effects of varying model parameters on the output. This will help us establish realistic bounds on some parameters, or even being able to fix other parameters. | Adequate |
| Other:__________ | ||||
| Additional Comments |
4. Limitations
| Disclaimer statement (explain key limitations) | Who needs to know about this disclaimer? | How is this disclaimer shared with that audience? | Current Conformance Level / Target Conformance Level |
|---|---|---|---|
| The current 3-state model that is used to generate contractile force in the model does not currently capture the influence of ATP. This will be added, with other states, in the near future. | Other cardiac modelers | This will be list at the begining of documentation for the code, as well as in publications. | Adequate |
| The current model only simulates the left ventricle. Thus the influence of the right ventricle on the deformation in the heart is not captured. | Other cardiac modelers | This will be list at the begining of documentation for the code, as well as in publications. | Adequate |
| Calcium transients are currently calculated with a simplified set of equations. This module is in the process of being enhanced. | Other cardiac modelers | This will be list at the begining of documentation for the code, as well as in publications. | Adequate |
5. Version control
| Current Conformance Level / Target Conformance Level |
|---|
| Extensive |
| Naming Conventions? | Repository? | Code Review? | |
|---|---|---|---|
| individual modeler | Yes, a common convention has been established and is used by all project participants. | All codes are version controlled on GitHub. | The code review process is hierarchical. Individual modelers review code with each other durin the implementation process. This work is overseen by the senior investigators and postdocs on the project. |
| within the lab | Yes, a common convention has been established and is used by all project participants. | All codes are version controlled on GitHub. | The code review process is hierarchical. Individual modelers review code with each other durin the implementation process. This work is overseen by the senior investigators and postdocs on the project. |
| collaborators | Yes, a common convention has been established and is used by all project participants. | All codes are version controlled on GitHub. | In addition to the description above for individual modelers and within the lab, there are regular video conference meetings to review code with collaborators at other insitutions. Since all members have access to the repository, code can be viewed anywhere. |
6. Documentation
| Current Conformance Level / Target Conformance Level | |
|---|---|
| Code commented? | Adequate. We are begining to use Doxygen for documentation generation. This allows for the generation of well organized documentation from code comments. This integrates all documentation from code to guides. |
| Scope and intended use described? | Adequate |
| User’s guide? | Partial |
| Developer’s guide? | Partial |
7. Dissemination
| Current Conformance Level / Target Conformance Level |
|---|
| Partial |
| Target Audience(s): | “Inner circle” | Scientific community | Public |
|---|---|---|---|
| Simulations | Are shared directly within the project team, across institutions. | Since the code is still being developed, the repository for sharing is still private. This will change once the code is functioning with all modules in place. | |
| Models | Are shared directly within the project team, across institutions. | Since the code is still being developed, the repository for sharing is still private. This will change once the code is functioning with all modules in place. | |
| Software | Is shared directly within the project team, across institutions. | Since the code is still being developed, the repository for sharing is still private. This will change once the code is functioning with all modules in place. | |
| Results | Are shared directly within the project team, across institutions. | Have been presented at several conferences (Biophys, SEM, etc). Results have also been published, for example: https://doi.org/10.3389/fphys.2018.00425 | |
| Implications of results | https://www.imagwiki.nibib.nih.gov/resources/publications/closing-therapeutic-loop | https://uknow.uky.edu/research/predicting-how-healthy-your-heart-will-be-years-down-road |
8. Independent reviews
| Current Conformance Level / Target Conformance Level |
|---|
| Adequate |
| Reviewer(s) name & affiliation: | Alberto Figueroa (University of Michigan) and Seungik Baek (Michigan State University) |
|---|---|
| When was review performed? | We plan to have two reviews per year. |
| How was review performed and outcomes of the review? | We have had initial planning meetings (via video conference). We plan to send each other test cases with our codes, in order for the each group to run the code and verify it works. We also plan to have onsight visits at the University of Kentucky and the University of Michigan to review the code and documentation. |
9. Test competing implementations
| Current Conformance Level / Target Conformance Level |
|---|
| Adequate |
| Yes or No (briefly summarize) | |
|---|---|
| Were competing implementations tested? | Yes. Test cases with the FEniCS framework have been compared to identical simulations in LS-Dyna. Also, single element simuations in FEniCS were compared to identical simulations in MyoSim. Both LS-Dyna and MyoSim are established codes. |
| Did this lead to model refinement or improvement? | Yes. We have improved the solution techniques used to solve the system of ODEs that govern the generation of contractile force. |
10. Conform to standards
| Current Conformance Level / Target Conformance Level |
|---|
| Adequate |
| Yes or No (briefly summarize) | |
|---|---|
| Are there operating procedures, guidelines, or standards for this type of multiscale modeling? | Yes. These are outlined nicely in the paper titled: "Considerations for Reporting Finite Element Analysis Studies in Biomechanics", by Erdemir et al. |
| How do your modeling efforts conform? | We conform to the standards ourlined in the above paper. Namely, we follow the standard format when describing model identification, model structure, simulation structure, verification, validation, and availability. |