Coronary Shear Stress Modeling

ISEF Category: Biomedical and Health Sciences

Subcategory: Pathophysiology  ·  Difficulty: Advanced  ·  Setup: University Lab  ·  Time: Full Year

The Hook

A plaque does not break open at random. Blood flow pushes on artery walls in uneven ways, and those force patterns can help reveal where a dangerous spot may form. That means you can study risk before a rupture ever happens. Your project can turn a medical problem into a testable flow problem.

What Is It?

This topic asks you to model how blood moves through a coronary bifurcation, which is where one artery splits into two. As blood turns through the split, it creates shear stress, the drag force of flowing blood on the vessel wall. Think of it like water rushing around a rock in a stream, where some spots get hit hard and others sit in slow, swirling water.

Researchers use that flow pattern to look for plaque sites that may be more likely to rupture. Plaque is a fatty buildup inside the artery wall, and rupture can trigger a heart attack. In this project, you would build a computer model from public vessel images, run the flow simulation, and compare the predicted high-risk regions with follow-up imaging from TCIA.

Why This Is a Good Topic

This is a strong science fair topic because you can test a clear cause-and-effect question with public data and free software. You are not guessing from a picture, you are measuring flow, comparing models, and checking whether the same spots stay risky across cases. It connects directly to heart disease, which keeps the project real. You can also learn image segmentation, CFD basics, and data analysis without needing a wet lab.

Research Questions

• How does branch angle change peak wall shear stress at coronary bifurcations? ?
• What is the effect of plaque location on oscillatory shear index in the parent vessel and side branch? ?
• Does low wall shear stress predict the same rupture-prone site across patient-derived geometries? ?
• To what extent do steady-flow and pulsatile-flow assumptions change the rank order of risky sites? ?
• Which geometric feature, branch angle, curvature, or lumen narrowing, best matches later TCIA imaging findings? ?
• How does mesh refinement near the bifurcation change the location of the highest-stress zone? ?

Basic Materials

• A laptop or desktop computer with at least 16 GB of RAM.
• Free SimVascular installation.
• 3D Slicer for image review and segmentation.
• Public coronary imaging datasets from TCIA or a similar open repository.
• Python with NumPy, pandas, SciPy, and Matplotlib.
• Spreadsheet software for tracking cases, model settings, and output metrics.

Advanced Materials

• A workstation with a dedicated GPU for large meshes.
• Access to institutional imaging archives or a mentor-provided dataset with follow-up scans.
• Mesh generation and quality-check tools such as Gmsh or MeshLab.
• Python with advanced statistics packages or R for deeper analysis.
• Version control with Git and GitHub for code, geometry files, and result tracking.
• A high-performance computing cluster for running multiple patient cases.

Software & Tools

• SimVascular: Builds vessel geometries and runs blood-flow simulations.
• 3D Slicer: Segments coronary images and checks anatomy before meshing.
• ParaView: Visualizes flow fields, wall shear stress maps, and bifurcation hotspots.
• Python: Cleans simulation output and compares hemodynamic metrics across cases.
• ImageJ: Measures image features when you need a quick check against segmented scans.

Experiment Steps

1. Define the exact clinical question you want to test, then choose one hemodynamic metric to treat as your main outcome.
2. Select a small set of public coronary geometries that share the same type of bifurcation, so your comparison stays fair.
3. Build a consistent segmentation and meshing plan, then decide how you will check that geometry changes do not come from bad image processing.
4. Set boundary-condition rules before running simulations, so every case follows the same flow assumptions.
5. Plan a validation method that compares predicted high-risk sites with follow-up imaging, not just with a single time point.
6. Decide which statistics will tell you whether geometry, plaque position, or flow assumptions matter most.

Common Pitfalls

• Mixing scans from different cardiac phases, which shifts the vessel shape and changes the stress map.
• Comparing patients without normalizing flow boundary conditions, which can make one geometry look risky only because the input flow was larger.
• Using a mesh that is too coarse at the carina, which hides the stress spike where plaque often matters most.
• Treating low wall shear stress as the only risk signal, which can miss oscillatory flow and recirculation near the branch.
• Calling a model validated after one image match, which ignores whether the predicted hotspot repeats across patients and time points.

What Makes This Competitive

A strong version of this project goes beyond making a colorful flow map. You would compare at least two hemodynamic metrics, test more than one modeling assumption, and check whether the same prediction holds across several patient geometries. A sharper entry also asks which geometry features matter most, not just whether stress is high. That kind of analysis shows real scientific judgment.

Project Variations

• Compare left main bifurcations with LAD or circumflex bifurcations to see whether branch type changes the stress pattern.
• Swap steady flow for pulsatile flow and test whether the rupture-prone site stays in the same place.
• Add a geometry-only analysis that compares branch angle, curvature, and narrowing against the hemodynamic result.

Learn More

• SimVascular Documentation: Learn vessel modeling and CFD workflows in the official tutorials and user guides.
• The Cancer Imaging Archive (TCIA): Find public imaging datasets and follow-up scans by searching the TCIA collections.
• PubMed: Search review articles on coronary bifurcation hemodynamics, wall shear stress, and plaque rupture.
• 3D Slicer Documentation: Follow free tutorials for image segmentation and anatomy review.
• NHLBI: Read plain-language background on coronary artery disease and plaque rupture on the National Heart, Lung, and Blood Institute site.
• MIT OpenCourseWare Fluid Mechanics: Review core flow concepts before you build your model.