Early Lung Cancer Diagnosis from Virtual Follow-up LDCT Generation via Correlational Autoencoder and Latent Flow Matching

2025-11-22

AI Summary

Early lung cancer diagnosis faces challenges in differentiating benign from malignant pulmonary nodules and suffers from diagnostic delays due to reliance on serial follow-up CT scans. Existing AI methods typically operate on single-scan data and fail to model nodule dynamics over time. To address this, we propose CorrFlowNet—the first generative model integrating a correlation-aware autoencoder with latent-space flow matching, augmented by neural ordinary differential equations (ODEs) and an auxiliary classifier. It synthesizes high-fidelity virtual one-year follow-up low-dose CT images from baseline scans, explicitly modeling nodule progression patterns. Crucially, CorrFlowNet quantifies malignancy risk without requiring real follow-up scans. Evaluated on clinical data, it achieves diagnostic accuracy comparable to actual follow-up CT (AUC improvement of 8.2%), significantly shortening diagnostic turnaround time. This work establishes a novel, interpretable, and deployable paradigm for temporal medical image generation in lung cancer screening.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Natural Language Processing: Generation

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Lung cancer is one of the most commonly diagnosed cancers, and early diagnosis is critical because the survival rate declines sharply once the disease progresses to advanced stages. However, achieving an early diagnosis remains challenging, particularly in distinguishing subtle early signals of malignancy from those of benign conditions. In clinical practice, a patient with a high risk may need to undergo an initial baseline and several annual follow-up examinations (e.g., CT scans) before receiving a definitive diagnosis, which can result in missing the optimal treatment. Recently, Artificial Intelligence (AI) methods have been increasingly used for early diagnosis of lung cancer, but most existing algorithms focus on radiomic features extraction from single early-stage CT scans. Inspired by recent advances in diffusion models for image generation, this paper proposes a generative method, named CorrFlowNet, which creates a virtual, one-year follow-up CT scan after the initial baseline scan. This virtual follow-up would allow for an early detection of malignant/benign nodules, reducing the need to wait for clinical follow-ups. During training, our approach employs a correlational autoencoder to encode both early baseline and follow-up CT images into a latent space that captures the dynamics of nodule progression as well as the correlations between them, followed by a flow matching algorithm on the latent space with a neural ordinary differential equation. An auxiliary classifier is used to further enhance the diagnostic accuracy. Evaluations on a real clinical dataset show our method can significantly improve downstream lung nodule risk assessment compared with existing baseline models. Moreover, its diagnostic accuracy is comparable with real clinical CT follow-ups, highlighting its potential to improve cancer diagnosis.

Problem

• Generating virtual follow-up CT scans from baseline scans for early lung cancer diagnosis
• Distinguishing malignant from benign nodules without waiting for clinical follow-ups
• Improving lung nodule risk assessment accuracy using AI-generated CT progression

Innovation

• Generates virtual follow-up CT scans using correlational autoencoder
• Employs latent flow matching with neural ODEs
• Enhances diagnosis via auxiliary classifier for accuracy