BrainPuzzle: Hybrid Physics and Data-Driven Reconstruction for Transcranial Ultrasound Tomography

2025-10-22

AI Summary

Transcranial ultrasound brain imaging suffers from severe acoustic attenuation, mode conversion, and phase distortion induced by the skull, compounded by limited spatial coverage and low signal-to-noise ratio due to clinically constrained small-aperture transducers. Conventional physics-based methods lack quantitative accuracy, while purely data-driven approaches fail to model the nonlinear, nonlocal wave propagation in bone, leading to systematic quantitative bias. This paper proposes a physics-informed deep learning framework comprising two stages: (1) multi-angle time-reversal migration for initial image formation; and (2) a Transformer-based super-resolution encoder-decoder network augmented with graph attention mechanisms to jointly optimize sound-speed map reconstruction under sparse-aperture conditions. Evaluated on synthetic data, the method significantly improves quantitative accuracy and anatomical fidelity compared to purely physics-based or purely data-driven baselines, advancing the clinical translation of quantitative transcranial ultrasound neuroimaging.

Technology Category

• Computer Vision: Low Level & Physics-based Vision
• Machine Learning: Multimodal Learning
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Security and Privacy: Data transparency and provenance

Abstract

Ultrasound brain imaging remains challenging due to the large difference in sound speed between the skull and brain tissues and the difficulty of coupling large probes to the skull. This work aims to achieve quantitative transcranial ultrasound by reconstructing an accurate speed-of-sound (SoS) map of the brain. Traditional physics-based full-waveform inversion (FWI) is limited by weak signals caused by skull-induced attenuation, mode conversion, and phase aberration, as well as incomplete spatial coverage since full-aperture arrays are clinically impractical. In contrast, purely data-driven methods that learn directly from raw ultrasound data often fail to model the complex nonlinear and nonlocal wave propagation through bone, leading to anatomically plausible but quantitatively biased SoS maps under low signal-to-noise and sparse-aperture conditions. To address these issues, we propose BrainPuzzle, a hybrid two-stage framework that combines physical modeling with machine learning. In the first stage, reverse time migration (time-reversal acoustics) is applied to multi-angle acquisitions to produce migration fragments that preserve structural details even under low SNR. In the second stage, a transformer-based super-resolution encoder-decoder with a graph-based attention unit (GAU) fuses these fragments into a coherent and quantitatively accurate SoS image. A partial-array acquisition strategy using a movable low-count transducer set improves feasibility and coupling, while the hybrid algorithm compensates for the missing aperture. Experiments on two synthetic datasets show that BrainPuzzle achieves superior SoS reconstruction accuracy and image completeness, demonstrating its potential for advancing quantitative ultrasound brain imaging.

Problem

• Reconstructing accurate brain speed-of-sound maps through skull
• Overcoming limitations of traditional physics-based ultrasound reconstruction methods
• Addressing sparse-aperture and low-SNR challenges in transcranial imaging

Innovation

• Hybrid physics and data-driven reconstruction framework
• Transformer-based encoder-decoder with graph attention unit
• Partial-array acquisition with movable transducers