DiffSOS: Acoustic Conditional Diffusion Model for Speed-of-Sound Reconstruction in Ultrasound Computed Tomography

2026-02-27

AI Summary

This work proposes DiffSOS, a physics-constrained conditional diffusion model for ultrasound computed tomography (USCT) that addresses the longstanding trade-off between computational efficiency and reconstruction fidelity in sound speed imaging. By integrating an acoustic ControlNet to rigorously couple wavefield measurements and designing a hybrid loss function that jointly optimizes noise prediction, spatial reconstruction, and spectral content, DiffSOS achieves high-fidelity reconstructions in just 10 DDIM sampling steps. The method further enables pixel-level uncertainty quantification to assess result reliability. Evaluated on the OpenPros USCT benchmark, DiffSOS attains a multi-scale structural similarity index of 0.957, substantially outperforming existing approaches and marking the first demonstration of near-real-time, high-fidelity sound speed reconstruction.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Constraint Satisfaction and Optimization: Constraint Optimization
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Accurate Speed-of-Sound (SoS) reconstruction from acoustic waveforms is a cornerstone of ultrasound computed tomography (USCT), enabling quantitative velocity mapping that reveals subtle anatomical details and pathological variations often invisible in conventional imaging. However, practical utility is hindered by the limitations of existing algorithms; traditional Full Waveform Inversion (FWI) is computationally intensive, while current deep learning approaches tend to produce oversmoothed results lacking fine details. We propose DiffSOS, a conditional diffusion model that directly maps acoustic waveforms to SoS maps. Our framework employs a specialized acoustic ControlNet to strictly ground the denoising process in physical wave measurements. To ensure structural consistency, we optimize a hybrid loss function that integrates noise prediction, spatial reconstruction, and noise frequency content. To accelerate inference, we employ stochastic Denoising Diffusion Implicit Model (DDIM) sampling, achieving near real-time reconstruction with only 10 steps. Crucially, we exploit the stochastic generative nature of our framework to estimate pixel-wise uncertainty, providing a measure of reliability that is often absent in deterministic approaches. Evaluated on the OpenPros USCT benchmark, DiffSOS significantly outperforms state-of-the-art networks, achieving an average Multi-scale Structural Similarity of 0.957. Our approach provides high-fidelity SoS maps with a principled measure of confidence, facilitating safer and faster clinical interpretation.

Problem

• Speed-of-Sound reconstruction
• Ultrasound Computed Tomography
• Acoustic Waveforms
• Quantitative Imaging
• Image Detail Preservation

Innovation

• conditional diffusion model
• acoustic ControlNet
• Speed-of-Sound reconstruction
• uncertainty quantification
• ultrasound computed tomography