Masked cross-domain self-supervised deep learning framework for photoacoustic computed tomography reconstruction

被引:3
作者
Lan, Hengrong [1 ]
Huang, Lijie [1 ]
Wei, Xingyue [1 ]
Li, Zhiqiang [1 ]
Lv, Jing [2 ]
Ma, Cheng [3 ]
Nie, Liming [2 ]
Luo, Jianwen [1 ]
机构
[1] Tsinghua Univ, Sch Biomed Engn, Beijing 100084, Peoples R China
[2] Southern Med Univ, Guangdong Prov Peoples Hosp, Med Res Inst, Guangdong Acad Med Sci, Guangzhou 510080, Peoples R China
[3] Tsinghua Univ, Dept Elect Engn, Beijing 100084, Peoples R China
基金
中国国家自然科学基金; 中国博士后科学基金;
关键词
Inverse problem; Self-supervision; Deep learning; Photoacoustic; Transformer; IMAGE-RECONSTRUCTION; ALGORITHM;
D O I
10.1016/j.neunet.2024.106515
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Accurate image reconstruction is crucial for photoacoustic (PA) computed tomography (PACT). Recently, deep learning has been used to reconstruct PA images with a supervised scheme, which requires high-quality images as ground truth labels. However, practical implementations encounter inevitable trade-offs between cost and performance due to the expensive nature of employing additional channels for accessing more measurements. Here, we propose a masked cross-domain self-supervised (CDSS) reconstruction strategy to overcome the lack of ground truth labels from limited PA measurements. We implement the self-supervised reconstruction in a modelbased form. Simultaneously, we take advantage of self-supervision to enforce the consistency of measurements and images across three partitions of the measured PA data, achieved by randomly masking different channels. Our findings indicate that dynamically masking a substantial proportion of channels, such as 80%, yields meaningful self-supervisors in both the image and signal domains. Consequently, this approach reduces the multiplicity of pseudo solutions and enables efficient image reconstruction using fewer PA measurements, ultimately minimizing reconstruction error. Experimental results on in-vivo PACT dataset of mice demonstrate the potential of our self-supervised framework. Moreover, our method exhibits impressive performance, achieving a structural similarity index (SSIM) of 0.87 in an extreme sparse case utilizing only 13 channels, which outperforms the performance of the supervised scheme with 16 channels (0.77 SSIM). Adding to its advantages, our method can be deployed on different trainable models in an end-to-end manner, further enhancing its versatility and applicability.
引用
收藏
页数:12
相关论文
共 54 条
[31]   Multiscale optical and optoacoustic imaging of amyloid-β deposits in mice [J].
Ni, Ruiqing ;
Chen, Zhenyue ;
Dean-Ben, Xose Luis ;
Voigt, Fabian F. ;
Kirschenbaum, Daniel ;
Shi, Gloria ;
Villois, Alessia ;
Zhou, Quanyu ;
Crimi, Alessandro ;
Arosio, Paolo ;
Nitsch, Roger M. ;
Nilsson, K. Peter R. ;
Aguzzi, Adriano ;
Helmchen, Fritjof ;
Klohs, Jan ;
Razansky, Daniel .
NATURE BIOMEDICAL ENGINEERING, 2022, 6 (09) :1031-+
[32]   Image reconstruction of the absorption coefficients with l1-norm minimization from photoacoustic measurements\ [J].
Okawa, Shinpei ;
Hirasawa, Takeshi ;
Kushibiki, Toshihiro ;
Ishihara, Miya .
QUANTITATIVE IMAGING IN MEDICINE AND SURGERY, 2015, 5 (01) :78-85
[33]   Photoacoustic Reconstruction Using Sparsity in Curvelet Frame: Image Versus Data Domain [J].
Pan, Bolin ;
Arridge, Simon R. ;
Lucka, Felix ;
Cox, Ben T. ;
Huynh, Nam ;
Beard, Paul C. ;
Zhang, Edward Z. ;
Betcke, Marta M. .
IEEE TRANSACTIONS ON COMPUTATIONAL IMAGING, 2021, 7 :879-893
[34]   Fractional Regularization to Improve Photoacoustic Tomographic Image Reconstruction [J].
Prakash, Jaya ;
Sanny, Dween ;
Kalva, Sandeep Kumar ;
Pramanik, Manojit ;
Yalavarthy, Phaneendra K. .
IEEE TRANSACTIONS ON MEDICAL IMAGING, 2019, 38 (08) :1935-1947
[35]  
Reiter A., 2017, SPIE, P504
[36]   U-Net: Convolutional Networks for Biomedical Image Segmentation [J].
Ronneberger, Olaf ;
Fischer, Philipp ;
Brox, Thomas .
MEDICAL IMAGE COMPUTING AND COMPUTER-ASSISTED INTERVENTION, PT III, 2015, 9351 :234-241
[37]  
Si C., 2022, arXiv
[38]  
Singh M. K. A., 2020, SPIE, P161
[39]   Sparse-view reconstruction for photoacoustic tomography combining diffusion model with model-based iteration [J].
Song, Xianlin ;
Wang, Guijun ;
Zhong, Wenhua ;
Guo, Kangjun ;
Li, Zilong ;
Liu, Xuan ;
Dong, Jiaqing ;
Liu, Qiegen .
PHOTOACOUSTICS, 2023, 33
[40]   Photoacoustic clinical imaging [J].
Steinberg, Idan ;
Huland, David M. ;
Vermesh, Ophir ;
Frostig, Hadas E. ;
Tummers, Willemieke S. ;
Gambhir, Sanjiv S. .
PHOTOACOUSTICS, 2019, 14 :77-98