Score-based Generative Models for Photoacoustic Image Reconstruction with Rotation Consistency Constraints

TL;DR

This paper introduces RCC-SGM, a novel score-based generative model with rotation consistency constraints, improving photoacoustic image reconstruction from limited sensor data by leveraging probabilistic sampling.

Contribution

The paper presents a new score-based generative model that incorporates rotation consistency constraints, enabling better generalization across measurement processes in photoacoustic imaging.

Findings

Achieves higher PSNR (32.29) with limited measurements compared to supervised methods.

Generalizes well to different measurement processes without retraining.

Utilizes iterative Langevin dynamics with rotation constraints for image reconstruction.

Abstract

Photoacoustic tomography (PAT) is a newly emerged imaging modality which enables both high optical contrast and acoustic depth of penetration. Reconstructing images of photoacoustic tomography from limited amount of senser data is among one of the major challenges in photoacoustic imaging. Previous works based on deep learning were trained in supervised fashion, which directly map the input partially known sensor data to the ground truth reconstructed from full field of view. Recently, score-based generative models played an increasingly significant role in generative modeling. Leveraging this probabilistic model, we proposed Rotation Consistency Constrained Score-based Generative Model (RCC-SGM), which recovers the PAT images by iterative sampling between Langevin dynamics and a constraint term utilizing the rotation consistency between the images and the measurements. Our proposed method can generalize to different measurement processes (32.29 PSNR with 16 measurements under random sampling, whereas 28.50 for supervised counterpart), while supervised methods need to train on specific inverse mappings.