Abstract
Predicting final ischemic infarct volumes from acute imaging is a cornerstone of personalized stroke management, yet current strategies remain polarized between uninterpretable machine learning architectures and overly detailed electrophysiological models that are intractable in clinical imaging settings. We bridge this clinical gap by introducing the first imaging-driven framework that parameterizes a Fisher-KPP reaction-diffusion partial differential equation (PDE) directly from clinical MRI. The continuous state variable $u(\mathbf{x},t)\in[0,1]$ models tissue damage, capturing the forward expansion of ionic stress through the extracellular space alongside a localized metabolic commitment to cell death gated within the baseline perfusion deficit ($T_{\max}>6$\,s). We evaluate this paradigm as an oracle: model and threshold parameters are fit to the 90-day outcome, so the results describe an upper bound on the model’s feasibility and potential. On a subset of the ISLES 2017 dataset ($N=29$), incorporating biophysical propagation constraints yields a mean Dice score of $0.46 \pm 0.24$, compared to $0.25 \pm 0.21$ for standard rCBF thresholding. Ablations show that spatially varying the diffusion field with clinical perfusion maps improves on a reaction-only baseline and captures penumbral expansion, while keeping the model fully interpretable. This proof-of-concept shows that first-principles physics can capture part of ischemic lesion evolution directly on clinical scan grids, a step toward patient-specific biophysical infarct forecasting.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/SWITCH_023.pdf
SharedIt Link: Not yet available
SpringerLink (DOI): Not yet available
Supplementary Material: Not Submitted
Link to Open Review
Open Review Page: https://openreview.net/forum?id=ZLQNnfWEi1
BibTex
@InProceedings{AbbMuh_An_MICCAISAT2026,
author = { Abbas, Muhammad Hussnain AND Balcerak, Michal AND Ahmad, Asif AND Menze, Bjoern AND de la Rosa, Ezequiel},
title = { { An Imaging-Informed Reaction-Diffusion Model of Infarct Growth } },
booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
year = {2026},
publisher = {Springer Nature Switzerland},
volume = {LNCS 17255},
month = {pending},
page = {pending}
}