Vascular Draining Confounds Laminar Decoding in fMRI

by Jonas Karolis Degutis, Denis Chaimow, Romy Lorenz

---

Overview

This repository contains the full simulation pipeline for exploring vascular draining confounds in laminar decoding with fMRI.
It provides scripts to reproduce the analyses and figures from the study.

---

How to Run

All simulations and analyses are implemented in Python. The main simulation functions are housed in simulation.py, while specific scripts reproduce the figures.

Scripts and Their Outputs

• runSimulationParallel.py

  • Simulates and plots data for:
    • Figure 1c (using layers=3)
    • Figure 2a (using layers=3)
    • Figure 5b (using layers=6)

• runSimulation_twoPatterns.py

  • Simulates and plots data for Figure 2b (using layers=3)
  • Simulates and plots data for Figure 5b for Deep and Superficial Different Patterns (using layers=6)

• runDeconvolution.py

  • Deconvolves simulated data and plots the results for Figure 1b and Figure 2b (using layers=3)
  • And Figure 5d (using layers=6)

• runSimulation_samePattern.py

  • Simulates and plots data for Figure 3 (using layers=3)
  • Simulates and plots data for Figure 6 (using layers=6)

• runVoxMisalignment.py and runVascModelMisalignment.py

  • Run analyses of misalignment effects, plotted in Figure 4

• simulation.py

  • Core simulation class containing the main functions used across all scripts.

---

Requirements

• Python 3.x
• Standard scientific Python stack (NumPy, SciPy, Matplotlib, Seaborn, Pandas, Scikit-lean, statsmodels)

(Found in requirements.txt)

---

Contact

For questions, please contact:
📧 j.karolis.degutis(at)gmail.com

---

License

MIT License

© J. Karolis Degutis, 2026