Continuous-Time Deep Glioma Growth Models

TL;DR

This paper introduces a continuous-time deep learning model for glioma growth prediction, capable of handling irregular data and producing realistic growth trajectories, which could enhance clinical decision-making.

Contribution

It extends Neural Processes with hierarchical multi-scale encoding and attention, enabling flexible, continuous-time tumor growth modeling from patient data.

Findings

Outperforms existing models on a 379-patient dataset

Captures global and local tumor growth variations

Produces consistent growth trajectories over continuous time

Abstract

The ability to estimate how a tumor might evolve in the future could have tremendous clinical benefits, from improved treatment decisions to better dose distribution in radiation therapy. Recent work has approached the glioma growth modeling problem via deep learning and variational inference, thus learning growth dynamics entirely from a real patient data distribution. So far, this approach was constrained to predefined image acquisition intervals and sequences of fixed length, which limits its applicability in more realistic scenarios. We overcome these limitations by extending Neural Processes, a class of conditional generative models for stochastic time series, with a hierarchical multi-scale representation encoding including a spatio-temporal attention mechanism. The result is a learned growth model that can be conditioned on an arbitrary number of observations, and that can produce a distribution of temporally consistent growth trajectories on a continuous time axis. On a dataset of 379 patients, the approach successfully captures both global and finer-grained variations in the images, exhibiting superior performance compared to other learned growth models.