SDEIT: Semantic-Driven Electrical Impedance Tomography

2025-04-05

AI Summary

EIT suffers from severe ill-posedness due to complex and variable anatomical structures. To address this, we propose a semantics-driven unsupervised reconstruction framework. For the first time, we integrate the large-scale text-to-image diffusion model Stable Diffusion 3.5 into EIT, leveraging natural language prompts as semantic priors. Our method couples implicit neural representations (INRs) with plug-and-play (PnP) optimization, enabling physics-constrained, high-fidelity reconstructions without paired training data. The core innovation lies in a tripartite prior fusion mechanism—textual, generative, and physical—which jointly enforces anatomical consistency and enhances fine structural recovery. Extensive experiments on both simulated and experimental datasets demonstrate that our approach consistently outperforms existing state-of-the-art methods. This validates the effectiveness and generalizability of multimodal semantic priors for solving ill-posed inverse problems in medical imaging.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Regularization methods using prior knowledge are essential in solving ill-posed inverse problems such as Electrical Impedance Tomography (EIT). However, designing effective regularization and integrating prior information into EIT remains challenging due to the complexity and variability of anatomical structures. In this work, we introduce SDEIT, a novel semantic-driven framework that integrates Stable Diffusion 3.5 into EIT, marking the first use of large-scale text-to-image generation models in EIT. SDEIT employs natural language prompts as semantic priors to guide the reconstruction process. By coupling an implicit neural representation (INR) network with a plug-and-play optimization scheme that leverages SD-generated images as generative priors, SDEIT improves structural consistency and recovers fine details. Importantly, this method does not rely on paired training datasets, increasing its adaptability to varied EIT scenarios. Extensive experiments on both simulated and experimental data demonstrate that SDEIT outperforms state-of-the-art techniques, offering superior accuracy and robustness. This work opens a new pathway for integrating multimodal priors into ill-posed inverse problems like EIT.

Problem

• Integrating semantic priors into EIT reconstruction
• Improving structural consistency without paired datasets
• Leveraging Stable Diffusion for generative priors

Innovation

• Integrates Stable Diffusion 3.5 into EIT
• Uses natural language prompts as semantic priors
• Couples INR network with plug-and-play optimization