# Cross-domain Self-supervised Framework for Photoacoustic Computed Tomography Image Reconstruction

arXiv: https://arxiv.org/abs/2301.06681 · HTML: https://mlsanity.com/paper/2301.06681/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 20 of its 69 references that are in the map.
- First posted on arXiv 17 Jan 2023.

## Builds on (most cited first, 20 of 20)

- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [Image reconstruction by domain transform manifold learning](https://mlsanity.com/paper/1704.08841/) (2017)
- [Inception Transformer](https://mlsanity.com/paper/2205.12956/) (2022)
- [NETT: Solving Inverse Problems with Deep Neural Networks](https://mlsanity.com/paper/1803.00092/) (2018)
- [Equivariant Imaging: Learning Beyond the Range Space](https://mlsanity.com/paper/2103.14756/) (2021)
- [Fully Dense UNet for 2D Sparse Photoacoustic Tomography Artifact Removal](https://mlsanity.com/paper/1808.10848/) (2018)
- [Model based learning for accelerated, limited-view 3D photoacoustic tomography](https://mlsanity.com/paper/1708.09832/) (2017)
- [Deep Learning in Photoacoustic Tomography: Current approaches and future directions](https://mlsanity.com/paper/2009.07608/) (2020)
- [Deep Learning for Photoacoustic Tomography from Sparse Data](https://mlsanity.com/paper/1704.04587/) (2017)
- [Deep Learning for Biomedical Image Reconstruction: A Survey](https://mlsanity.com/paper/2002.12351/) (2020)
- [Limited View and Sparse Photoacoustic Tomography for Neuroimaging with Deep Learning](https://mlsanity.com/paper/1911.04357/) (2019)
- [A Partially Learned Algorithm for Joint Photoacoustic Reconstruction and Segmentation](https://mlsanity.com/paper/1906.07499/) (2019)
- [Simultaneous reconstruction of the initial pressure and sound speed in photoacoustic tomography using a deep-learning approach](https://mlsanity.com/paper/1907.09951/) (2019)
- [Deep-learning Image Reconstruction for Real-time Photoacoustic System](https://mlsanity.com/paper/2001.04631/) (2020)
- [Sinogram super-resolution and denoising convolutional neural network (SRCN) for limited data photoacoustic tomography](https://mlsanity.com/paper/2001.06434/) (2020)
- [Deep Learning Improves Contrast in Low-Fluence Photoacoustic Imaging](https://mlsanity.com/paper/2004.08782/) (2020)
- [AS-Net: Fast Photoacoustic Reconstruction with Multi-feature Fusion from Sparse Data](https://mlsanity.com/paper/2101.08934/) (2021)

## Abstract (arXiv)

Accurate image reconstruction is crucial for photoacoustic (PA) computed tomography (PACT). Recently, deep learning has been used to reconstruct the PA image with a supervised scheme, which requires high-quality images as ground truth labels. In practice, there are inevitable trade-offs between cost and performance since the use of more channels is an expensive strategy to access more measurements. Here, we propose a cross-domain unsupervised reconstruction (CDUR) strategy with a pure transformer model, which overcomes the lack of ground truth labels from limited PA measurements. The proposed approach exploits the equivariance of PACT to achieve high performance with a smaller number of channels. We implement a self-supervised reconstruction in a model-based form. Meanwhile, we also leverage the self-supervision to enforce the measurement and image consistency on three partitions of measured PA data, by randomly masking different channels. We find that dynamically masking a high proportion of the channels, e.g., 80%, yields nontrivial self-supervisors in both image and signal domains, which decrease the multiplicity of the pseudo solution to efficiently reconstruct the image from fewer PA measurements with minimum error of the image. Experimental results on in-vivo PACT dataset of mice demonstrate the potential of our unsupervised framework. In addition, our method shows a high performance (0.83 structural similarity index (SSIM) in the extreme sparse case with 13 channels), which is close to that of supervised scheme (0.77 SSIM with 16 channels). On top of all the advantages, our method may be deployed on different trainable models in an end-to-end manner.