Dual-Domain Self-Supervised Artifact Removal Framework for Photoacoustic Computed Tomography

2026-07-14

AI Summary

This work addresses the severe artifacts in photoacoustic computed tomography under sparse-view acquisition by proposing a self-supervised artifact removal framework. The method uniquely leverages the complementary artifact characteristics between spatial and frequency domain reconstructions, introducing a twin lightweight network that operates without ground-truth labels. A composite loss function is designed to integrate cross-domain fidelity and uncertainty-weighted consistency, enabling effective disentanglement of dual-domain features and robust artifact suppression. Experimental results demonstrate that the proposed approach significantly reduces artifacts across simulated, phantom, in vivo rat, and human datasets while maintaining efficient end-to-end inference performance.

Abstract

Photoacoustic Computed Tomography (PACT) often faces severe challenges from reconstruction artifacts due to sparse detection conditions. In this work, based on the distinct differences in artifact patterns between back-projection-based and Fourier-based reconstruction algorithms, we propose a self-supervised artifact removal framework that employs a lightweight Siamese Neural Network and a composite loss function integrating cross-domain fidelity and uncertainty-weighted consistency, effectively decoupling dual-domain features and filtering artifacts. Comprehensive validations using simulations, phantoms, in vivo rat and human experimental data demonstrate that the proposed method can significantly suppress image artifacts. Furthermore, enabled by the acceleration of the spatial-domain and frequency-domain inverse operator, this end-to-end approach also achieves exceptional computational efficiency.

Problem

• Photoacoustic Computed Tomography
• reconstruction artifacts
• sparse detection
• artifact removal
• image quality

Innovation

• dual-domain
• self-supervised
• artifact removal
• Siamese Neural Network
• photoacoustic computed tomography