4D VQ-GAN: Synthesising Medical Scans at Any Time Point for Personalised Disease Progression Modelling of Idiopathic Pulmonary Fibrosis

2025-02-08

AI Summary

This study addresses personalized progression modeling for idiopathic pulmonary fibrosis (IPF) by proposing the first two-stage framework integrating 3D-VQ-GAN and neural ordinary differential equations (Neural ODEs) to enable end-to-end generation of high-fidelity 4D CT volumes at arbitrary time points and clinically interpretable survival prediction. Methodologically, 3D-VQ-GAN first learns discrete latent representations of longitudinal CT scans; subsequently, Neural ODEs model the continuous temporal evolution of these quantized embeddings, jointly optimized with imaging biomarker extraction and Cox regression for survival analysis. Experiments demonstrate that synthesized CT volumes achieve structural fidelity and pathological consistency comparable to real scans; the derived imaging biomarkers yield a C-index of 0.71 for survival prediction—on par with benchmarks using real CT data. This work pioneers a differentiable, unified temporal framework that bridges generative modeling and dynamic survival forecasting, establishing a novel paradigm for precision intervention in IPF.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Natural Language Processing: Generation

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Understanding the progression trajectories of diseases is crucial for early diagnosis and effective treatment planning. This is especially vital for life-threatening conditions such as Idiopathic Pulmonary Fibrosis (IPF), a chronic, progressive lung disease with a prognosis comparable to many cancers. Computed tomography (CT) imaging has been established as a reliable diagnostic tool for IPF. Accurately predicting future CT scans of early-stage IPF patients can aid in developing better treatment strategies, thereby improving survival outcomes. In this paper, we propose 4D Vector Quantised Generative Adversarial Networks (4D-VQ-GAN), a model capable of generating realistic CT volumes of IPF patients at any time point. The model is trained using a two-stage approach. In the first stage, a 3D-VQ-GAN is trained to reconstruct CT volumes. In the second stage, a Neural Ordinary Differential Equation (ODE) based temporal model is trained to capture the temporal dynamics of the quantised embeddings generated by the encoder in the first stage. We evaluate different configurations of our model for generating longitudinal CT scans and compare the results against ground truth data, both quantitatively and qualitatively. For validation, we conduct survival analysis using imaging biomarkers derived from generated CT scans and achieve a C-index comparable to that of biomarkers derived from the real CT scans. The survival analysis results demonstrate the potential clinical utility inherent to generated longitudinal CT scans, showing that they can reliably predict survival outcomes.

Problem

• Predicting future CT scans for IPF
• Modeling disease progression trajectories
• Generating realistic longitudinal medical images

Innovation

• 4D-VQ-GAN for CT synthesis
• Neural ODE captures dynamics
• Generates realistic IPF scans