TumorFlow: Physics-Guided Longitudinal MRI Synthesis of Glioblastoma Growth

TL;DR

This paper introduces TumorFlow, a physics-guided generative model that synthesizes realistic 3D MRI sequences of glioblastoma growth, enabling personalized tumor progression visualization and synthetic data generation for neuro-oncology.

Contribution

It presents a novel biophysically-conditioned generative framework combining tumor-infiltration maps with growth models for realistic, controllable tumor progression simulation.

Findings

Achieves 75% Dice overlap with biophysical model in longitudinal extrapolation.

Maintains PSNR of 25 in surrounding tissue across sequences.

Generates temporally coherent tumor growth trajectories.

Abstract

Glioblastoma exhibits diverse, infiltrative, and patient-specific growth patterns that are only partially visible on routine MRI, making it difficult to reliably assess true tumor extent and personalize treatment planning and follow-up. We present a biophysically-conditioned generative framework that synthesizes biologically realistic 3D brain MRI volumes from estimated, spatially continuous tumor-concentration fields. Our approach combines a generative model with tumor-infiltration maps that can be propagated through time using a biophysical growth model, enabling fine-grained control over tumor shape and growth while preserving patient anatomy. This enables us to synthesize consistent tumor growth trajectories directly in the space of real patients, providing interpretable, controllable estimation of tumor infiltration and progression beyond what is explicitly observed in imaging. We evaluate the framework on longitudinal glioblastoma cases and demonstrate that it can generate temporally coherent sequences with realistic changes in tumor appearance and surrounding tissue response. These results suggest that integrating mechanistic tumor growth priors with modern generative modeling can provide a practical tool for patient-specific progression visualization and for generating controlled synthetic data to support downstream neuro-oncology workflows. In longitudinal extrapolation, we achieve a consistent 75% Dice overlap with the biophysical model while maintaining a constant PSNR of 25 in the surrounding tissue. Our code is available at: https://github.com/valentin-biller/lgm.git