Fibroblast Alignment and Wound Closure Analysis

ISEF Category: Biomedical Engineering

Subcategory: Cell and Tissue Engineering  ·  Difficulty: Advanced  ·  Setup: University Lab  ·  Time: Full Year

The Hook

Your skin closes a wound by coordinating cells like a repair crew on a collapsing bridge. If those cells line up and move well, healing speeds up. This project lets you measure that pattern from real videos and ask what surface stiffness helps cells do that job best.

What Is It?

Fibroblasts are cells that help rebuild tissue after injury. In scratch-assay videos, scientists make a small gap in a cell layer and watch how the cells move to fill it. You can think of the gap as a tiny empty road and the cells as traffic closing it from both sides.

Alignment means the cells point in a similar direction. That matters because cells often move better when they organize as a group. Substrate stiffness is how stiff or soft the surface feels to the cells. Cells can sense that stiffness through a process called mechanosensing, which means they read physical cues and change behavior.

Your project combines two parts. First, you build a pipeline in ImageJ and, if you want, a simple machine-learning step to measure alignment in public scratch-assay videos. Second, you compare those results with published wound-healing data to look for stiffness ranges that match faster closure. That gives you both image analysis and biology in one project.

Why This Is a Good Topic

This is a strong science fair topic because you can measure real data instead of just describing biology. You can test a clear input, like alignment or stiffness, and a clear output, like wound-closure speed. The topic connects to wound repair, scar healing, tissue engineering, and how cells respond to their physical environment. You can also learn image analysis, data cleaning, statistics, and meta-analysis, which makes the project feel research-level without needing to grow your own cells.

Research Questions

• How does fibroblast alignment change as substrate stiffness increases across published scratch-assay videos? ?
• What is the effect of cell alignment on the rate of wound-gap closure in scratch assays? ?
• Does a machine-learning classifier improve alignment measurement compared with manual scoring in ImageJ? ?
• To what extent do published wound-healing kinetics cluster around a narrow stiffness range? ?
• Which alignment metric best predicts faster closure, orientation angle, aspect ratio, or coherence? ?
• How does video source quality affect the reliability of alignment measurements from public datasets? ?

Basic Materials

• Computer with internet access.
• ImageJ or Fiji software.
• Spreadsheet software such as Google Sheets or Excel.
• PubMed access through a school or public library.
• Publicly available scratch-assay videos or image sets.
• Notebook for tracking dataset IDs and metadata.
• Basic statistics calculator or spreadsheet functions.
• Optional, a simple ML notebook environment such as Google Colab.

Advanced Materials

• Computer with a dedicated GPU or strong CPU.
• Fiji with relevant plugins for particle analysis and orientation metrics.
• Python with NumPy, pandas, scikit-learn, and OpenCV.
• Jupyter Notebook or Google Colab.
• Access to full-text journal articles through a library.
• Image annotation tool for checking model labels.
• Reference manager such as Zotero.
• University library access for broader wound-healing literature.

Software & Tools

• ImageJ/Fiji: Measures cell orientation, edge movement, and image features from scratch-assay videos.
• Python: Handles cleaning, feature extraction, and model testing.
• Google Colab: Runs Python notebooks without local setup and helps with shared code.
• PubMed: Finds primary studies and review articles on fibroblast migration, stiffness, and wound healing.
• Zotero: Organizes papers and keeps your literature review organized.

Experiment Steps

1. Define the exact output you will measure, such as alignment angle, closure speed, or both.
2. Choose a public dataset set and write rules for which videos or images count.
3. Build one clean ImageJ workflow that measures alignment the same way every time.
4. Plan a small validation set so you can compare automated scores with manual scoring.
5. Organize published studies into a table with stiffness, cell type, assay type, and healing rate.
6. Decide how you will test whether alignment, stiffness, and closure speed move together.

Common Pitfalls

• Using videos with different microscopes or lighting, which makes alignment measurements hard to compare.
• Mixing cell types or assay formats, which hides any stiffness pattern you are trying to find.
• Letting the machine-learning model train on too few examples, which makes its predictions unstable.
• Comparing studies without standardizing units or time points, which turns the meta-analysis into noise.
• Measuring wound closure from frames with unclear edges, which creates false speed estimates.

What Makes This Competitive

A class-level version of this project stops at measuring a few videos. A stronger version builds a repeatable pipeline, checks it against human scoring, and tests whether the same trend appears across many studies. If you also separate effects from stiffness, cell type, and imaging quality, your analysis starts to look like real research. The best entries ask a sharper question than "does it work?" They ask when it works, for whom, and under what physical conditions.

Project Variations

• Compare fibroblast alignment in scratch assays from different species, such as human, mouse, or rat cells.
• Swap alignment for wound-edge migration speed and test whether the two metrics disagree in some datasets.
• Focus on one staining method or imaging style and test whether it changes how well ImageJ measures orientation.

Learn More

• PubMed: Search for review articles on fibroblast migration, mechanosensing, and wound-healing kinetics.
• NIH PubMed Central: Read free full-text papers on scratch assays and tissue repair methods.
• NIH RePORTER: Look up funded wound-healing and mechanobiology projects to see current research directions.
• MIT OpenCourseWare: Search for free materials on biological engineering, image analysis, and data science.
• Nature Reviews Molecular Cell Biology: Search for review articles on cell migration, mechanotransduction, and tissue repair through a library or public abstract access.