Convolutional Invasion and Expansion Networks for Tumor Growth Prediction

TL;DR

This paper introduces deep convolutional neural networks to directly model tumor cell invasion and mass-effect, enabling personalized and more accurate tumor growth predictions from multimodal imaging data.

Contribution

It proposes a novel deep learning framework that fuses invasion and expansion networks, outperforming traditional mathematical models in tumor growth prediction.

Findings

Outperforms state-of-the-art mathematical models in accuracy

Easily trained on population data and personalized to individual patients

Captures complementary information from invasion and expansion subnetworks

Abstract

Tumor growth is associated with cell invasion and mass-effect, which are traditionally formulated by mathematical models, namely reaction-diffusion equations and biomechanics. Such models can be personalized based on clinical measurements to build the predictive models for tumor growth. In this paper, we investigate the possibility of using deep convolutional neural networks (ConvNets) to directly represent and learn the cell invasion and mass-effect, and to predict the subsequent involvement regions of a tumor. The invasion network learns the cell invasion from information related to metabolic rate, cell density and tumor boundary derived from multimodal imaging data. The expansion network models the mass-effect from the growing motion of tumor mass. We also study different architectures that fuse the invasion and expansion networks, in order to exploit the inherent correlations among them. Our network can easily be trained on population data and personalized to a target patient, unlike most previous mathematical modeling methods that fail to incorporate population data. Quantitative experiments on a pancreatic tumor data set show that the proposed method substantially outperforms a state-of-the-art mathematical model-based approach in both accuracy and efficiency, and that the information captured by each of the two subnetworks are complementary.