ImageFlowNet: Forecasting Multiscale Image-Level Trajectories of Disease Progression with Irregularly-Sampled Longitudinal Medical Images

TL;DR

ImageFlowNet is a novel model that forecasts disease progression trajectories from longitudinal medical images by learning multiscale representations and flow fields, effectively handling irregular sampling and data sparsity.

Contribution

We introduce ImageFlowNet, a new model combining neural ODE/SDE frameworks with multiscale representations for disease trajectory forecasting from medical images.

Findings

Outperforms existing methods on three medical datasets.

Effectively handles irregular sampling and data sparsity.

Provides theoretical insights into neural ODE/SDE formulations.

Abstract

Advances in medical imaging technologies have enabled the collection of longitudinal images, which involve repeated scanning of the same patients over time, to monitor disease progression. However, predictive modeling of such data remains challenging due to high dimensionality, irregular sampling, and data sparsity. To address these issues, we propose ImageFlowNet, a novel model designed to forecast disease trajectories from initial images while preserving spatial details. ImageFlowNet first learns multiscale joint representation spaces across patients and time points, then optimizes deterministic or stochastic flow fields within these spaces using a position-parameterized neural ODE/SDE framework. The model leverages a UNet architecture to create robust multiscale representations and mitigates data scarcity by combining knowledge from all patients. We provide theoretical insights that support our formulation of ODEs, and motivate our regularizations involving high-level visual features, latent space organization, and trajectory smoothness. We validate ImageFlowNet on three longitudinal medical image datasets depicting progression in geographic atrophy, multiple sclerosis, and glioblastoma, demonstrating its ability to effectively forecast disease progression and outperform existing methods. Our contributions include the development of ImageFlowNet, its theoretical underpinnings, and empirical validation on real-world datasets. The official implementation is available at https://github.com/KrishnaswamyLab/ImageFlowNet.