Organoid Patterning With Reaction-Diffusion Models

ISEF Category: Biomedical Engineering

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

The Hook

A tiny change in chemical signals can decide where cells grow, cluster, or split apart. That makes patterning feel less like art and more like a code you can test. Your project asks whether a reaction-diffusion model can predict real organoid shapes and spacing. If it works, you have a bridge from math to tissue design.

What Is It?

Reaction-diffusion models describe how two or more chemicals spread and interact to make patterns. Think of a drop of ink in water, but with one substance speeding things up and another slowing them down. In the Gray-Scott or Turing style of model, that push and pull can create stripes, spots, or repeating bands.

For this project, you use that math to predict how tissue might organize during growth. Organoids are tiny lab-grown tissue models that can mimic parts of real organs. You are not growing a full organ here. You are testing whether a computer model can match the spacing, clustering, or symmetry seen in published organoid images. If the model fits well, it suggests a simple rule may help explain a complex biological pattern.

Why This Is a Good Topic

This is a strong science fair topic because you can ask clear, measurable questions. You can change model parameters, compare predicted patterns with real images, and measure how close they match. That gives you real data, not just pretty simulations. It also connects to tissue engineering, developmental biology, and regenerative medicine, so the work has real-world value. You can learn simulation, image analysis, and model validation without needing to invent a new wet-lab assay.

Research Questions

• How does changing the diffusion ratio affect the spacing between predicted tissue clusters?
• What is the effect of reaction strength on whether the model forms spots, stripes, or blobs?
• Does adding noise to the initial conditions change how closely the simulation matches organoid image patterns?
• To what extent do different parameter sets reproduce the digit-like spacing seen in published limb-bud datasets?
• Which pattern metrics, such as cluster count, nearest-neighbor distance, or symmetry, best separate good and poor model fits?
• How does grid resolution change the stability of the simulated pattern output?

Basic Materials

• Raspberry Pi or similar low-cost single-board computer with GPU support.
• Monitor, keyboard, and mouse.
• MicroSD card with operating system installed.
• Python installed with NumPy, SciPy, Matplotlib, and OpenCV.
• Public organoid image datasets from journal supplements, NIH, or PubMed-linked repositories.
• Spreadsheet software for recording parameter sets and outputs.
• Digital caliper or image-analysis workflow for measuring distances in saved images.
• External storage or cloud backup for simulation files.

Advanced Materials

• University workstation or GPU server for faster parameter sweeps.
• Image segmentation software for tissue boundary extraction.
• Python libraries for scientific computing, optimization, and image comparison.
• Access to a curated organoid image dataset with metadata.
• Fiji or ImageJ for measuring pattern spacing and area.
• Statistical software for model fitting and comparison.
• Version control system such as Git for tracking code changes.
• Optional microscopy images from a tissue engineering lab for external validation.

Software & Tools

• Python: Runs the reaction-diffusion simulation and stores parameter outputs for later analysis.
• NumPy: Handles arrays for the simulation grid and numerical updates.
• Matplotlib: Plots pattern formation and compares runs across parameter settings.
• ImageJ: Measures spacing, area, and symmetry in organoid images.
• OpenCV: Helps segment images and extract pattern features from published datasets.

Experiment Steps

1. Define the biological pattern you want to match, such as cluster spacing, boundary shape, or repeat count.
2. Choose the simplest reaction-diffusion model that can produce that pattern and justify why it fits the biology.
3. Build a parameter sweep plan so you can test many settings without changing your code by hand.
4. Decide how you will compare simulated patterns with published images using one or more quantitative metrics.
5. Plan controls that separate true pattern matching from accidental visual similarity.
6. Set a rule for selecting your best model fit, then test whether it still works on a second dataset.

Common Pitfalls

• Comparing simulations to publication figures without first correcting image scale, which makes spacing measurements meaningless.
• Treating a nice-looking pattern as a good fit even when the cluster count and symmetry do not match the real data.
• Changing several model parameters at once, which makes it impossible to tell which one caused the pattern shift.
• Using a single published image as proof, which makes the result fragile and easy to overfit.
• Ignoring segmentation error when extracting organoid boundaries, which can distort every downstream metric.

What Makes This Competitive

A strong project goes beyond making pretty patterns. You need a clear validation plan, a dataset you did not tune the model on, and a fair way to score prediction quality. A more competitive version compares several pattern metrics, not just one visual match. You can also test whether your model predicts one tissue type well, then fails on another, which helps you define where the model works and where it breaks.

Project Variations

• Compare Gray-Scott and classic Turing parameter sets on the same organoid image set to see which predicts spacing better.
• Swap limb-bud images for intestinal or neural organoid datasets and test whether one model generalizes across tissue types.
• Add shape analysis, such as fractal dimension or radial symmetry, to see whether the model captures more than just cluster spacing.

Learn More

• NIH PubMed: Search for review articles on reaction-diffusion models, organoids, and tissue patterning.
• NCBI Bookshelf: Read free textbook chapters on developmental biology and morphogen gradients.
• MIT OpenCourseWare: Look for computational biology, differential equations, or systems biology lectures that cover modeling ideas.
• ImageJ documentation: Learn how to measure distances, areas, and segmentation outputs in biological images.
• Nature Reviews or Development articles: Search for review papers on organoid patterning and limb development through your school or public library access.