Deep Image Prior for photoacoustic tomography can mitigate limited-view artifacts

TL;DR

This paper demonstrates that the deep image prior (DIP) framework effectively reconstructs photoacoustic tomography images, reducing artifacts and noise in limited-view scenarios without supervision.

Contribution

It introduces an unsupervised DIP-based approach for PAT reconstruction, incorporating fast algorithms and regularization techniques, outperforming classical TV methods.

Findings

DIP improves image quality over traditional TV reconstructions.

The method is effective with simulated and experimental PAT data.

DIP provides robust reconstructions in limited-view geometries.

Abstract

We study the deep image prior (DIP) framework applied to photoacoustic tomography (PAT) as an unsupervised reconstruction approach to mitigate limited-view artifacts and noise commonly encountered in experimental settings. Efficient implementation is achieved by employing recently published fast forward and adjoint algorithms for circular measurement geometries. Initialization via a fast inverse and total variation (TV) regularization are applied to further suppress noise and mitigate overfitting. For comparison, we compute a classical TV reconstruction. Our experiments comprise simulated PAT measurements under limited-view geometries and varying levels of added noise as well as experimental measurements together with using a digital twin for quality assessment. Our findings suggest that DIP framework provides an effective unsupervised strategy for robust PAT reconstruction even in the challenging case of a limited view geometry providing improvement in several quantitative measures over total variation reconstructions.