Deep Probabilistic Modeling of Glioma Growth
This work addresses the challenge of predicting brain tumor progression for medical imaging and treatment planning, offering a novel method that could improve accuracy over traditional models.
The authors tackled the problem of modeling glioma growth by proposing a data-driven approach that learns tumor growth dynamics implicitly from data, without relying on explicit biological models, and demonstrated its ability to learn a distribution of plausible future tumor appearances based on past observations.
Existing approaches to modeling the dynamics of brain tumor growth, specifically glioma, employ biologically inspired models of cell diffusion, using image data to estimate the associated parameters. In this work, we propose an alternative approach based on recent advances in probabilistic segmentation and representation learning that implicitly learns growth dynamics directly from data without an underlying explicit model. We present evidence that our approach is able to learn a distribution of plausible future tumor appearances conditioned on past observations of the same tumor.