Efficient Flow Matching for Sparse-View CT Reconstruction

2026-02-27

AI Summary

This work addresses the inefficiency and data consistency conflicts arising from the stochastic nature of diffusion models in sparse-view CT reconstruction by proposing FMCT, a deterministic reconstruction framework based on Flow Matching (FM), along with its efficient variant EFMCT. By leveraging the inherently deterministic ODE trajectories generated through FM, the method naturally aligns with data consistency constraints. Furthermore, it introduces an innovative velocity field reuse mechanism that substantially reduces the number of neural network evaluations. Under the guarantee of bounded reconstruction error, the proposed approach achieves image quality comparable to that of diffusion models while significantly lowering the number of function evaluations (NFEs), thereby markedly improving inference efficiency.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Reasoning under Uncertainty: Stochastic Optimization
• Machine Learning: Feature Construction/Reformulation

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Web Mining and Content Analysis: Content-based information diffusion

Abstract

Generative models, particularly Diffusion Models (DM), have shown strong potential for Computed Tomography (CT) reconstruction serving as expressive priors for solving ill-posed inverse problems. However, diffusion-based reconstruction relies on Stochastic Differential Equations (SDEs) for forward diffusion and reverse denoising, where such stochasticity can interfere with repeated data consistency corrections in CT reconstruction. Since CT reconstruction is often time-critical in clinical and interventional scenarios, improving reconstruction efficiency is essential. In contrast, Flow Matching (FM) models sampling as a deterministic Ordinary Differential Equation (ODE), yielding smooth trajectories without stochastic noise injection. This deterministic formulation is naturally compatible with repeated data consistency operations. Furthermore, we observe that FM-predicted velocity fields exhibit strong correlations across adjacent steps. Motivated by this, we propose an FM-based CT reconstruction framework (FMCT) and an efficient variant (EFMCT) that reuses previously predicted velocity fields over consecutive steps to substantially reduce the number of Neural network Function Evaluations (NFEs), thereby improving inference efficiency. We provide theoretical analysis showing that the error introduced by velocity reuse is bounded when combined with data consistency operations. Extensive experiments demonstrate that FMCT/EFMCT achieve competitive reconstruction quality while significantly improving computational efficiency compared with diffusion-based methods. The codebase is open-sourced at https://github.com/EFMCT/EFMCT.

Problem

• Sparse-View CT Reconstruction
• Diffusion Models
• Flow Matching
• Computational Efficiency
• Data Consistency

Innovation

• Flow Matching
• Sparse-View CT Reconstruction
• Deterministic ODE
• Velocity Field Reuse
• Efficient Inference