A Partially-Learned Algorithm for Joint Photo-acoustic Reconstruction and Segmentation

Boink, Yoeri E.; Manohar, Srirang; Brüne, Christoph

doi:10.1109/tmi.2019.2922026

Cited by 64 publications

(45 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, we will explore the possibility to perform photoacoustic reconstruction and segmentation simultaneously using a partially learned algorithm. 20 As a basis for this future work, the robustness analysis performed here demonstrates that it is important to have enough variety in image intensities in the training set.…”

Section: Discussionmentioning

confidence: 95%

Robustness of a partially learned photoacoustic reconstruction algorithm

Boink

Brüne

Manohar

2019

Photons Plus Ultrasound: Imaging and Sensing 2019

Self Cite

View full text Add to dashboard Cite

Classical reconstruction algorithms for photoacoustic tomography (PAT) are mathematically proven to converge, but can be very slow and inadequate with respect to model and data assumptions. With the help of neural networks, learned reconstruction algorithms have recently been developed. These learned algorithms have shown to surpass the reconstruction quality of non-learned ones, but their mathematical analysis is challenging, therefore convergence and stability are not guaranteed. In this work, we combine the structure of a well-known model-based algorithm with the efficiency of a datadriven neural network. We show its robustness in simulation against uncertainty and changes in PAT system settings. This is done by varying the placement and calibration of detectors, as well as varying properties of the synthetic phantom, such as geometrical structure and intensity. This yields a robust and computationally efficient algorithm that outperforms classical methods and can be applied to the PAT problem in a universal setting.

show abstract

Section: Discussionmentioning

confidence: 95%

Robustness of a partially learned photoacoustic reconstruction algorithm

Boink

Brüne

Manohar

2019

Photons Plus Ultrasound: Imaging and Sensing 2019

Self Cite

View full text Add to dashboard Cite

show abstract

“…For reconstructions in PAT, the work by Boink et al [137][138][139] has demonstrated the robustness of these learned iterative schemes to a number of in silico phantoms as well as in an experimental study. The authors consider an extension to the learned gradient schemes introduced above called learned primal dual 18 (LPD) based on the successful primal-dual hybrid gradient method 140 (also known as the Chambolle-Pock algorithm).…”

Section: Learned Primal Dual In 2dmentioning

confidence: 99%

“…5(b). In their work, [137][138][139] the authors examined the robustness of LPD with respect to changes in the target, including the contrast, background, structural changes, and noise level. They found that if the network is trained only on the basic training data, it generalizes fairly well with respect to noise (1 dB degradation in PSNR) and structural changes (3 dB), but is most sensitive to changes in background (7 dB) and contrast (11 dB).…”

Section: Learned Primal Dual In 2dmentioning

confidence: 99%

“…They found that if the network is trained only on the basic training data, it generalizes fairly well with respect to noise (1 dB degradation in PSNR) and structural changes (3 dB), but is most sensitive to changes in background (7 dB) and contrast (11 dB). 138 Additionally, the authors combine their learned reconstruction with a joint segmentation that is learned with the same network as an additional output and is shown to provide increased robustness compared to a reconstruction by filtered backprojection and segmentation with U-Net.…”

Section: Learned Primal Dual In 2dmentioning

confidence: 99%

See 1 more Smart Citation

Deep learning in photoacoustic tomography: current approaches and future directions

2020

View full text Add to dashboard Cite

Biomedical photoacoustic tomography, which can provide high-resolution 3D soft tissue images based on optical absorption, has advanced to the stage at which translation from the laboratory to clinical settings is becoming possible. The need for rapid image formation and the practical restrictions on data acquisition that arise from the constraints of a clinical workflow are presenting new image reconstruction challenges. There are many classical approaches to image reconstruction, but ameliorating the effects of incomplete or imperfect data through the incorporation of accurate priors is challenging and leads to slow algorithms. Recently, the application of deep learning (DL), or deep neural networks, to this problem has received a great deal of attention. We review the literature on learned image reconstruction, summarizing the current trends and explain how these approaches fit within, and to some extent have arisen from, a framework that encompasses classical reconstruction methods. In particular, it shows how these techniques can be understood from a Bayesian perspective, providing useful insights. We also provide a concise tutorial demonstration of three prototypical approaches to learned image reconstruction. The code and data sets for these demonstrations are available to researchers. It is anticipated that it is in in vivo applications-where data may be sparse, fast imaging critical, and priors difficult to construct by hand-that DL will have the most impact. With this in mind, we conclude with some indications of possible future research directions. © The Authors. Published by SPIE under a Creative Commons Attribution 4.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

show abstract

“…They designed a DNN to represent the iteration framework and introduced the gradient information of the photoacoustic data (Figure 7B). Boink et al designed a CNN network based on partially learned algorithm to simultaneously achieve image reconstruction and segmentation of PAT (96). Compared with the strategy of using U-Net for post-processing network, the reconstruction results are further improved by the iterative network.…”

Section: Patmentioning

confidence: 99%

A review of the application of machine learning in molecular imaging

Yin¹,

Cao²,

Wang³

et al. 2021

Ann Transl Med

View full text Add to dashboard Cite

Molecular imaging (MI) is a science that uses imaging methods to reflect the changes of molecular level in living state and conduct qualitative and quantitative studies on its biological behaviors in imaging. Optical molecular imaging (OMI) and nuclear medical imaging are two key research fields of MI.OMI technology refers to the optical information generated by the imaging target (such as tumors) due to drug intervention and other reasons. By collecting the optical information, researchers can track the motion trajectory of the imaging target at the molecular level. Owing to its high specificity and sensitivity, OMI has been widely used in preclinical research and clinical surgery. Nuclear medical imaging mainly detects ionizing radiation emitted by radioactive substances. It can provide molecular information for early diagnosis, effective treatment and basic research of diseases, which has become one of the frontiers and hot topics in the field of medicine in the world today. Both OMI and nuclear medical imaging technology require a lot of data processing and analysis. In recent years, artificial intelligence technology, especially neural networkbased machine learning (ML) technology, has been widely used in MI because of its powerful data processing capability. It provides a feasible strategy to deal with large and complex data for the requirement of MI. In this review, we will focus on the applications of ML methods in OMI and nuclear medical imaging.

show abstract

A Partially-Learned Algorithm for Joint Photo-acoustic Reconstruction and Segmentation

Cited by 64 publications

References 60 publications

Robustness of a partially learned photoacoustic reconstruction algorithm

Robustness of a partially learned photoacoustic reconstruction algorithm

Deep learning in photoacoustic tomography: current approaches and future directions

A review of the application of machine learning in molecular imaging

Contact Info

Product

Resources

About