MDiff-FMT: Morphology-aware Diffusion Model for Fluorescence Molecular Tomography with Small-scale Datasets

2024-10-09

AI Summary

FMT inverse problems are highly ill-posed, and existing deep learning methods suffer from poor morphological reconstruction fidelity and strong data dependency under limited training samples. To address this, we propose the first morphology-aware diffusion model for FMT reconstruction: a conditional denoising diffusion probabilistic model (DDPM) that leverages fluorescence images as structural priors; explicitly models the morphological degradation process via noise evolution to enable progressive, high-fidelity structural recovery; and incorporates a morphology-guided sampling strategy alongside a small-data training protocol. This work pioneers the integration of diffusion models into FMT reconstruction. In both numerical simulations and phantom experiments, our method achieves state-of-the-art performance—significantly improving boundary sharpness and anatomical consistency—while requiring no large-scale annotated datasets.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Search and Optimization: Sampling/Simulation-based Search

Application Category

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

Abstract

Fluorescence molecular tomography (FMT) is a sensitive optical imaging technology widely used in biomedical research. However, the ill-posedness of the inverse problem poses a huge challenge to FMT reconstruction. Although end-to-end deep learning algorithms have been widely used to address this critical issue, they still suffer from high data dependency and poor morphological restoration. In this paper, we report for the first time a morphology-aware diffusion model, MDiff-FMT, based on denoising diffusion probabilistic model (DDPM) to achieve high-fidelity morphological reconstruction for FMT. First, we use the noise addition of DDPM to simulate the process of the gradual degradation of morphological features, and achieve fine-grained reconstruction of morphological features through a stepwise probabilistic sampling mechanism, avoiding problems such as loss of structure details that may occur in end-to-end deep learning methods. Additionally, we introduce the conditional fluorescence image as structural prior information to sample a high-fidelity reconstructed image from the noisy images. Numerous numerical and real phantom experimental results show that the proposed MDiff-FMT achieves SOTA results in morphological reconstruction of FMT without relying on large-scale datasets.

Problem

• Addresses FMT reconstruction challenges with small datasets
• Improves morphological feature restoration in FMT imaging
• Reduces data dependency in deep learning-based FMT reconstruction

Innovation

• Morphology-aware diffusion model for FMT
• Denoising diffusion probabilistic model (DDPM)
• Conditional fluorescence image as prior