Diagnostics

您所在的位置：网站首页 › journal of magnetic resonance image › Diagnostics

Diagnostics

2023-12-25 08:14| 来源: 网络整理| 查看: 265

Next Article in Journal Comparison of Ultrasound Scores in Differentiating between Benign and Malignant Adnexal Masses Next Article in Special Issue A Semi-Supervised Graph Convolutional Network for Early Prediction of Motor Abnormalities in Very Preterm Infants Previous Article in Journal “Quo Vadis Diagnosis”: Application of Informatics in Early Detection of Pneumothorax Previous Article in Special Issue Efficient U-Net Architecture with Multiple Encoders and Attention Mechanism Decoders for Brain Tumor Segmentation Journals Active Journals Find a Journal Proceedings Series Topics Information For Authors For Reviewers For Editors For Librarians For Publishers For Societies For Conference Organizers Open Access Policy Institutional Open Access Program Special Issues Guidelines Editorial Process Research and Publication Ethics Article Processing Charges Awards Testimonials Author Services Initiatives Sciforum MDPI Books Preprints.org Scilit SciProfiles Encyclopedia JAMS Proceedings Series About Overview Contact Careers News Press Blog Sign In / Sign Up Notice clear Notice

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

Continue Cancel clear

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess.

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Journals Active Journals Find a Journal Proceedings Series Topics Information For Authors For Reviewers For Editors For Librarians For Publishers For Societies For Conference Organizers Open Access Policy Institutional Open Access Program Special Issues Guidelines Editorial Process Research and Publication Ethics Article Processing Charges Awards Testimonials Author Services Initiatives Sciforum MDPI Books Preprints.org Scilit SciProfiles Encyclopedia JAMS Proceedings Series About Overview Contact Careers News Press Blog Sign In / Sign Up Submit Journals Diagnostics Volume 13 Issue 7 10.3390/diagnostics13071306

Submit to this Journal Review for this Journal Propose a Special Issue ► ▼ Article Menu Article Menu Academic Editor

Nikolaos Dikaios Subscribe SciFeed Recommended Articles Related Info Links PubMed/Medline Google Scholar More by Authors Links on DOAJ Hossain, M. Biddut Kwon, K. Shinde, R. Kiran Imtiaz, S. Md Kim, N. on Google Scholar Hossain, M. Biddut Kwon, K. Shinde, R. Kiran Imtiaz, S. Md Kim, N. on PubMed Hossain, M. Biddut Kwon, K. Shinde, R. Kiran Imtiaz, S. Md Kim, N. /ajax/scifeed/subscribe Article Views Citations - Table of Contents Altmetric share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up ... Endorse textsms ... Comment Need Help? Support

Find support for a specific problem in the support section of our website.

Get Support Feedback

Please let us know what you think of our products and services.

Give Feedback Information

Visit our dedicated information section to learn more about MDPI.

Get Information clear JSmol Viewer clear first_page Download PDF settings Order Article Reprints Font Type: Arial Georgia Verdana Font Size: Aa Aa Aa Line Spacing:    Column Width:    Background: Open AccessArticle A Hybrid Residual Attention Convolutional Neural Network for Compressed Sensing Magnetic Resonance Image Reconstruction by

Md. Biddut Hossain

Md. Biddut Hossain Scilit Preprints.org Google Scholar View Publications ORCID

Ki-Chul Kwon

Ki-Chul Kwon Scilit Preprints.org Google Scholar View Publications ORCID

Rupali Kiran Shinde

Rupali Kiran Shinde Scilit Preprints.org Google Scholar View Publications ORCID

Shariar Md Imtiaz

Shariar Md Imtiaz Scilit Preprints.org Google Scholar View Publications ORCID

and

Nam Kim

Nam Kim Scilit Preprints.org Google Scholar View Publications * ORCID

School of Information and Communication Engineering, Chungbuk National University, Cheongju-Si 28644, Chungcheongbuk-Do, Republic of Korea * Author to whom correspondence should be addressed. Diagnostics 2023, 13(7), 1306; https://doi.org/10.3390/diagnostics13071306 Submission received: 20 February 2023 / Revised: 20 March 2023 / Accepted: 29 March 2023 / Published: 30 March 2023 (This article belongs to the Special Issue Artificial Intelligence for Magnetic Resonance Imaging) Download keyboard_arrow_down Download PDF Download PDF with Cover Download XML Download Epub Browse Figures Versions Notes

Abstract: We propose a dual-domain deep learning technique for accelerating compressed sensing magnetic resonance image reconstruction. An advanced convolutional neural network with residual connectivity and an attention mechanism was developed for frequency and image domains. First, the sensor domain subnetwork estimates the unmeasured frequencies of k-space to reduce aliasing artifacts. Second, the image domain subnetwork performs a pixel-wise operation to remove blur and noisy artifacts. The skip connections efficiently concatenate the feature maps to alleviate the vanishing gradient problem. An attention gate in each decoder layer enhances network generalizability and speeds up image reconstruction by eliminating irrelevant activations. The proposed technique reconstructs real-valued clinical images from sparsely sampled k-spaces that are identical to the reference images. The performance of this novel approach was compared with state-of-the-art direct mapping, single-domain, and multi-domain methods. With acceleration factors (AFs) of 4 and 5, our method improved the mean peak signal-to-noise ratio (PSNR) to 8.67 and 9.23, respectively, compared with the single-domain Unet model; similarly, our approach increased the average PSNR to 3.72 and 4.61, respectively, compared with the multi-domain W-net. Remarkably, using an AF of 6, it enhanced the PSNR by 9.87 ± 1.55 and 6.60 ± 0.38 compared with Unet and W-net, respectively. Keywords: MRI reconstruction; attention mechanism; compressed sensing; data augmentation; deep learning; dual residual CNN 1. IntroductionMagnetic resonance imaging (MRI) is a noninvasive and sophisticated medical imaging method that sheds light on the anatomical structure and operation of the human body and brain [1] by generating high-quality images. Each tissue’s unique properties can be recognized using a novel MRI acquisition method. Based on the different tissue signal diversities, MRI reconstructs quantitative images that are very important for the early diagnosis of sickness or physical changes. MRI does not entail exposure to harmful radiation [2], unlike X-rays, photoacoustic tomography, and computed tomography. However, the long image acquisition time [3] makes it challenging to use MRI in time-sensitive situations, such as in cases of stroke, although acquisition time can be reduced given that MRI systems allow for comprehensive control of data acquisition. Acquired frequencies in MRI are stored in k-space instead of image space. K-space is a matrix the same size as the reconstructed image that stores complex (real and imaginary) raw MRI data. Every point in this matrix holds a portion of the data needed to create the entire image. The periphery of k-space possesses high spatial frequency that depicts information concerning image edges, details, and sharp transitions. On the other hand, the central area of k-space retains the high spatial frequency that expresses the image at its brightest. Fully sampled k-space is essential for obtaining high-resolution images but increases acquisition time. Acquiring a few frequencies is one of the most popular methodologies for rapid MRI reconstruction. However, due to undersampling, tissue structures are often distorted, and aliasing artifacts appear in the images. Compressed sensing (CS) [4] randomly employs an iterative process to select appropriate frequencies for reconstructing suitable images from sparsely sampled MRI data. However, these iterative methods are time-consuming, which makes them challenging to use in conjunction with fast MRI.Deep learning (DL) has been effectively applied for the analysis of medical images [5,6]. Deep neural networks have emerged in medical image reconstruction, classification, and computer-based disease identification [7]. The early detection of a tumor is more important for effective treatment, especially when seeking to avoid surgery and reduce the risk of death. DL in computer-aided diagnosis (CAD) systems increases the accuracy and efficiency of diagnosing the normal (non-tumor zone) and abnormal (tumor zone) tissues at different stages that can be used in smart healthcare systems. Recently, DL has been used to address the shortcomings of iterative conventional CS techniques. DL is crucial for the efficient generation of high-quality images from undersampled k-spaces and obviates the need for repeated processing after the model has been appropriately trained by providing it with both fully and partially sampled k-spaces from corresponding images. A neural network [8] is first utilized to lessen aliasing artifacts in multi-coil MRI [9]. DL-based repetitive unrolled optimization approaches [10,11,12] can then be applied for CS-MRI reconstruction to better learn image features; however, the processing times for the iteration are relatively long and the ill-posed inverse problem remains to be solved. Several strategies [13,14,15,16,17,18,19,20,21,22] have been used to enhance the quality of distorted images. Distorted images can be reconstructed from sparse spatial frequencies using an inverse fast Fourier transform (IFFT); however, the visual characteristics may be restored incorrectly when frequencies are significantly sparse. Frequency domain networks [23,24,25,26] estimate the unknown missing k-space frequencies before image conversion; these networks also tend to recover low-importance sensor data that increase their reconstruction time. Direct mapping networks [27,28] straightforwardly convert the frequency into an image; however, they only work for relatively small images ( 1 s per image reconstruction. The proposed RA-CNN generates better images within an average of 0.6 s. Therefore, computation time and cost are reduced. Notably, the results showed that the AG-based methods performed better than the other single and cascade networks at higher AFs. Even though we are currently just testing our method with brain data and different sampling rates, not with other types of MRI datasets involving areas such as the knee or abdomen, the results are still significant. 6. ConclusionsThe proposed dual-domain RA-CNN reconstructs MRI images from sparsely sampled k-space data using two neural networks. The first CNN in the sensor domain predicts unacquired frequencies and then applies the second CNN in the image domain for image enhancement. Furthermore, each network has a unique impact on MRI reconstruction. Edge content and geometry are restored more effectively from undersampled k-space using this multi-domain CNN. As a CNN is used directly to retrieve the sensor data, some lower frequencies might be recoverable. Consequently, this method is capable of extracting realistic visual features and reconstructing images that are identical to real images. Since the visual characteristics are preserved, radiologists can interpret data accurately and rapidly. Residual connection significantly enhances feature reuse and network data flow. Moreover, AGs mix lower and higher spatial data to identify valuable features while using fewer parameters than the other sophisticated Unet-based methods. Although network training takes a long time, images can be generated rapidly after training.We show that the aggregation of two domains has an impact on MRI reconstruction performance. In end-to-end reconstruction based on residual and attention mechanisms, the RA-CNN performed better than several alternative single- and multi-domain networks, as reflected in the PSNR and SSIM values under various sampling rates. In future research, we will apply our strategy for interactive temperature-based MRI reconstruction for real-time diagnostics and therapy. Author ContributionsConceptualization, M.B.H.; data curation, M.B.H.; formal analysis, M.B.H.; funding acquisition, N.K.; investigation, M.B.H.; methodology, M.B.H.; project administration, K.-C.K. and N.K.; resources, M.B.H.; software, M.B.H.; supervision, N.K.; validation, M.B.H. and K.-C.K.; visualization, M.B.H., R.K.S. and S.M.I.; writing—original draft, M.B.H.; writing—review and editing, M.B.H., K.-C.K. and R.K.S. All authors have read and agreed to the published version of the manuscript.FundingThis work was assisted by the Institute for Information and Communications Technology Promotion (IITP) and was financed by the Korean government (MSIP) (No. 2021-0-00490; Development of precision analysis and imaging technology for biological radio waves) and a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (NRF-NRF- 2020R1A2C1101258).Institutional Review Board StatementNot applicable.Informed Consent StatementNot applicable.Data Availability StatementThe dataset is accessible from: https://sites.google.com/view/calgary-campinas-dataset/home (accessed on 20 January 2023). The source code of this manuscript is available at https://github.com/biddut2j8/RA-CNN (last accessed on 20 March 2023).AcknowledgmentsWe thank Anuja Anil Padwal, Ashwini Rural Collage, Solapur-413006, India, for validating the reconstructed images from a medical perspective. Further, the authors thank Shahinur Alam for checking the proposed method from a deep-learning perspective.Conflicts of InterestThe authors declare no conflict of interest. Moreover, Anuja Padwal has no intention of gaining financial or other advantages from this method.ReferencesBrown, R.W.; Cheng, Y.-C.N.; Haacke, E.M.; Thompson, M.R.; Venkatesan, R. Magnetic Resonance Imaging: Physical Principles and Sequence Design, 2nd ed.; John Wiley & Sons Ltd: Chichester, UK, 2014; ISBN 9781118633953. [Google Scholar]Cercignani, M.; Dowell, N.G.; Tofts, P.S. Quantitative MRI of the Brain: Principles of Physical Measurement; Cercignani, M., Nicholas, G., Dowell, P.S.T., Eds.; CRC Press: Boca Raton, FL, USA, 2018; ISBN 9781315363578. [Google Scholar]Muckley, M.J.; Riemenschneider, B.; Radmanesh, A.; Kim, S.; Jeong, G.; Ko, J.; Jun, Y.; Shin, H.; Hwang, D.; Mostapha, M.; et al. Results of the 2020 fastMRI challenge for machine learning MR image reconstruction. IEEE Trans. Med. Imaging 2021, 40, 2306–2317. [Google Scholar] [CrossRef]Lustig, M.; Donoho, D.; Pauly, J.M. Sparse MRI: The application of compressed sensing for rapid MR imaging. Magn. Reson. Med. 2007, 58, 1182–1195. [Google Scholar] [CrossRef] [PubMed]Lee, J.-G.; Jun, S.; Cho, Y.-W.; Lee, H.; Kim, G.B.; Seo, J.B.; Kim, N. Deep learning in medical imaging: General overview. Korean J. Radiol. 2017, 18, 570–584. [Google Scholar] [CrossRef] [PubMed][Green Version]Ahishakiye, E.; Van Gijzen, M.B.; Tumwiine, J.; Wario, R.; Obungoloch, J. A survey on deep learning in medical image reconstruction. Intell. Med. 2021, 1, 118–127. [Google Scholar] [CrossRef]Ji, L.; Mao, R.; Wu, J.; Ge, C.; Xiao, F.; Xu, X.; Xie, L.; Gu, X. Deep convolutional neural network for nasopharyngeal carcinoma discrimination on mri by comparison of hierarchical and simple layered convolutional neural networks. Diagnostics 2022, 12, 2478. [Google Scholar] [CrossRef]Kwon, K.; Kim, D.; Park, H. A parallel MR imaging method using multilayer perceptron. Med. Phys. 2017, 44, 6209–6224. [Google Scholar] [CrossRef][Green Version]Deshmane, A.; Gulani, V.; Griswold, M.A.; Seiberlich, N. Parallel MR imaging. J. Magn. Reson. Imaging 2012, 36, 55–72. [Google Scholar] [CrossRef][Green Version]Qin, C.; Schlemper, J.; Caballero, J.; Price, A.N.; Hajnal, J.V.; Rueckert, D. Convolutional recurrent neural networks for dynamic MR image reconstruction. IEEE Trans. Med. Imaging 2019, 38, 280–290. [Google Scholar] [CrossRef][Green Version]Zhang, X.; Lian, Q.; Yang, Y.; Su, Y. A deep unrolling network inspired by total variation for compressed sensing MRI. Digit. Signal Process. 2020, 107, 102856. [Google Scholar] [CrossRef]Hosseini, S.A.H.; Yaman, B.; Moeller, S.; Hong, M.; Akcakaya, M. Dense recurrent neural networks for accelerated MRI: History-cognizant unrolling of optimization algorithms. IEEE J. Sel. Top. Signal Process. 2020, 14, 1280–1291. [Google Scholar] [CrossRef]Yang, G.; Yu, S.; Dong, H.; Slabaugh, G.; Dragotti, P.L.; Ye, X.; Liu, F.; Arridge, S.; Keegan, J.; Guo, Y.; et al. DAGAN: Deep de-aliasing generative adversarial networks for fast compressed sensing MRI reconstruction. IEEE Trans. Med. Imaging 2018, 37, 1310–1321. [Google Scholar] [CrossRef] [PubMed][Green Version]Quan, T.M.; Nguyen-Duc, T.; Jeong, W.-K. Compressed sensing MRI reconstruction using a generative adversarial network with a cyclic loss. IEEE Trans. Med. Imaging 2018, 37, 1488–1497. [Google Scholar] [CrossRef] [PubMed][Green Version]Yuan, Z.; Jiang, M.; Wang, Y.; Wei, B.; Li, Y.; Wang, P.; Menpes-Smith, W.; Niu, Z.; Yang, G. SARA-GAN: Self-attention and relative average discriminator based generative adversarial networks for fast compressed sensing MRI reconstruction. Front. Neuroinform. 2020, 14, 611666. [Google Scholar] [CrossRef]Nath, R.; Callahan, S.; Singam, N.; Stoddard, M.; Amini, A.A. Accelerated phase contrast magnetic resonance imaging via deep learning. In Proceedings of the 2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI), Iowa City, IA, USA, 3–7 April 2020; IEEE: Piscatway, NJ, USA, 2020; Volume 3, pp. 834–838. [Google Scholar]Chen, Y.; Christodoulou, A.G.; Zhou, Z.; Shi, F.; Xie, Y.; Li, D. MRI super-resolution with GAN and 3D multi-level densenet: Smaller, faster, and better. arXiv 2020, arXiv:2003.01217. [Google Scholar]Jiang, M.; Zhi, M.; Wei, L.; Yang, X.; Zhang, J.; Li, Y.; Wang, P.; Huang, J.; Yang, G. FA-GAN: Fused attentive generative adversarial networks for MRI image super-resolution. Comput. Med. Imaging Graph. 2021, 92, 101969. [Google Scholar] [CrossRef]Zhang, K.; Hu, H.; Philbrick, K.; Conte, G.M.; Sobek, J.D.; Rouzrokh, P.; Erickson, B.J. SOUP-GAN: Super-resolution MRI using generative adversarial networks. Tomography 2022, 8, 905–919. [Google Scholar] [CrossRef]Li, W.; Feng, X.; An, H.; Ng, X.Y.; Zhang, Y.-J. MRI reconstruction with interpretable pixel-wise operations using reinforcement learning. Proc. AAAI Conf. Artif. Intell. 2020, 34, 792–799. [Google Scholar] [CrossRef]Andrew, J.; Mhatesh, T.S.R.; Sebastin, R.D.; Sagayam, K.M.; Eunice, J.; Pomplun, M.; Dang, H. Super-resolution reconstruction of brain magnetic resonance images via lightweight autoencoder. Informatics Med. Unlocked 2021, 26, 100713. [Google Scholar] [CrossRef]Wang, Y.; Pang, Y.; Tong, C. DSMENet: Detail and structure mutually enhancing network for under-sampled MRI reconstruction. Comput. Biol. Med. 2022, 154, 106204. [Google Scholar] [CrossRef]Han, Y.; Sunwoo, L.; Ye, J.C. K-space deep learning for accelerated MRI. IEEE Trans. Med. Imaging 2020, 39, 377–386. [Google Scholar] [CrossRef][Green Version]Pineda, L.; Basu, S.; Romero, A.; Calandra, R.; Drozdzal, M. Active MR k-space sampling with reinforcement learning. In Medical Image Computing and Computer Assisted Intervention–MICCAI 2020: 23rd International Conference, Lima, Peru, October 4–8, 2020; Proceedings, Part II 23; Springer International Publishing: Berlin/Heidelberg, Germany, 2020; pp. 23–33. [Google Scholar]Du, T.; Zhang, H.; Li, Y.; Pickup, S.; Rosen, M.; Zhou, R.; Song, H.K.; Fan, Y. Adaptive convolutional neural networks for accelerating magnetic resonance imaging via k-space data interpolation. Med. Image Anal. 2021, 72, 102098. [Google Scholar] [CrossRef] [PubMed]Arefeen, Y.; Beker, O.; Cho, J.; Yu, H.; Adalsteinsson, E.; Bilgic, B. Scan-specific artifact reduction in k-space (SPARK) neural networks synergize with physics-based reconstruction to accelerate MRI. Magn. Reson. Med. 2022, 87, 764–780. [Google Scholar] [CrossRef]Oh, C.; Kim, D.; Chung, J.-Y.; Han, Y.; Park, H. ETER-net: End to end MR image reconstruction using recurrent neural network. In Proceedings of the Lecture Notes in Computer Science Book Series (Volume 11074), International Workshop on Machine Learning for Medical Image Reconstruction, Granada, Spain, 16 September 2018; Springer International Publishing: Granada, Spain, 2018; pp. 12–20. [Google Scholar]Schlemper, J.; Oksuz, I.; Clough, J.; Duan, J.; King, A.P.; Schnabel, J.A.; Hajnal, J.V.; Rueckert, D. dAUTOMAP: Decomposing AUTOMAP to achieve scalability and enhance performance. arXiv 2019, arXiv:1909.10995. [Google Scholar]Eo, T.; Jun, Y.; Kim, T.; Jang, J.; Lee, H.J.; Hwang, D. KIKI-net: Cross-domain convolutional neural networks for reconstructing undersampled magnetic resonance images. Magn. Reson. Med. 2018, 80, 2188–2201. [Google Scholar] [CrossRef]Souza, R.; Lebel, R.M.; Frayne, R. A hybrid, dual domain, cascade of convolutional neural networks for magnetic resonance image reconstruction. Proc. Mach. Learn. Res. 2019, 102, 437–446. [Google Scholar]Ke, Z.; Zhu, Y.; Liang, D. Cascaded residual dense networks for dynamic MR imaging with edge-enhanced loss constraint. Investig. Magn. Reson. Imaging 2020, 24, 214. [Google Scholar] [CrossRef]Wang, Z.; Jiang, H.; Du, H.; Xu, J.; Qiu, B. IKWI-net: A cross-domain convolutional neural network for undersampled magnetic resonance image reconstruction. Magn. Reson. Imaging 2020, 73, 1–10. [Google Scholar] [CrossRef]Zhang, Y.; Lyu, J.; Bi, X. A dual-task dual-domain model for blind MRI reconstruction. Comput. Med. Imaging Graph. 2021, 89, 101862. [Google Scholar] [CrossRef] [PubMed]Schlemper, J.; Caballero, J.; Hajnal, J.V.; Price, A.N.; Rueckert, D. A deep cascade of convolutional neural networks for dynamic MR image reconstruction. IEEE Trans. Med. Imaging 2018, 37, 491–503. [Google Scholar] [CrossRef] [PubMed][Green Version]Souza, R.; Frayne, R. A hybrid frequency-domain/image-domain deep network for magnetic resonance image reconstruction. In Proceedings of the 2019 32nd SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI), Rio de Janeiro, Brazil, 28–30 October 2019; IEEE: Piscatway, NJ, USA, 2019; pp. 257–264. [Google Scholar]Ronneberger, O.; Fischer, P.; Brox, T. U-net: Convolutional networks for biomedical image segmentation. In Proceedings of the IMedical Image Computing and Computer-Assisted Intervention—MICCAI 2015. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Munich, Germany, 5–9 October 2015; Springer: Cham, Germany, 2015; Volume 9351, pp. 234–241. [Google Scholar]Shafiq, M.; Gu, Z. Deep residual learning for image recognition: A survey. Appl. Sci. 2022, 12, 8972. [Google Scholar] [CrossRef]Aghabiglou, A.; Eksioglu, E.M. MR image reconstruction using densely connected residual convolutional networks. Comput. Biol. Med. 2021, 139, 105010. [Google Scholar] [CrossRef]Por, E.; van Kooten, M.; Sarkovic, V. Nyquist–Shannon Sampling Theorem; Leiden University: Leiden, The Netherlands, 2019; p. 1. [Google Scholar]Sanz, J.; Huang, T. Discrete and continuous band-limited signal extrapolation. IEEE Trans. Acoust. 1983, 31, 1276–1285. [Google Scholar] [CrossRef]Haldar, J.P. Low-rank modeling of local k-space neighborhoods (LORAKS) for constrained MRI. IEEE Trans. Med. Imaging 2014, 33, 668–681. [Google Scholar] [CrossRef] [PubMed][Green Version]Oktay, O.; Schlemper, J.; Folgoc, L.L.; Lee, M.; Heinrich, M.; Misawa, K.; Mori, K.; McDonagh, S.; Hammerla, N.Y.; Kainz, B.; et al. Attention U-net: Learning where to look for the pancreas. arXiv 2018, arXiv:1804.03999. [Google Scholar]Santurkar, S.; Tsipras, D.; Ilyas, A.; Madry, A. How does batch normalization help optimization? Adv. Neural Inf. Process. Syst. 2018, 31, 2483–2493. [Google Scholar]Agarap, A.F. Deep learning using rectified linear units (ReLU). arXiv 2018, arXiv:1803.08375. [Google Scholar]Schlemper, J.; Oktay, O.; Schaap, M.; Heinrich, M.; Kainz, B.; Glocker, B.; Rueckert, D. Attention gated networks: Learning to leverage salient regions in medical images. Med. Image Anal. 2019, 53, 197–207. [Google Scholar] [CrossRef]Souza, R.; Lucena, O.; Garrafa, J.; Gobbi, D.; Saluzzi, M.; Appenzeller, S.; Rittner, L.; Frayne, R.; Lotufo, R. An open, multi-vendor, multi-field-strength brain MR dataset and analysis of publicly available skull stripping methods agreement. Neuroimage 2018, 170, 482–494. [Google Scholar] [CrossRef]Calgary-Campinas Public Brain MR Dataset. Available online: https://sites.google.com/view/calgary-campinas-dataset/home (accessed on 20 January 2023).Chlap, P.; Min, H.; Vandenberg, N.; Dowling, J.; Holloway, L.; Haworth, A. A review of medical image data augmentation techniques for deep learning applications. J. Med. Imaging Radiat. Oncol. 2021, 65, 545–563. [Google Scholar] [CrossRef]Fabian, Z.; Heckel, R.; Soltanolkotabi, M. Data augmentation for deep learning based accelerated MRI reconstruction with limited data. In Proceedings of the International Conference on Machine Learning, Vienna, Austria, 18–24 July 2021. [Google Scholar]Wang, Z.; Bovik, A.C.; Sheikh, H.R.; Simoncelli, E.P. Image quality assessment: From error visibility to structural similarity. IEEE Trans. Image Process. 2004, 13, 600–612. [Google Scholar] [CrossRef][Green Version]Aghabiglou, A.; Eksioglu, E.M. Projection-based cascaded U-net model for MR image reconstruction. Comput. Methods Programs Biomed. 2021, 207, 106151. [Google Scholar] [CrossRef] [PubMed]Hossain, M.B.; Kwon, K.-C.; Imtiaz, S.M.; Nam, O.-S.; Jeon, S.-H.; Kim, N. De-aliasing and accelerated sparse magnetic resonance image reconstruction using fully dense CNN with attention gates. Bioengineering 2022, 10, 22. [Google Scholar] [CrossRef] [PubMed]Jethi, A.K.; Murugesan, B.; Ram, K.; Sivaprakasam, M. Dual-encoder-Unet for fast MRI reconstruction. In Proceedings of the 2020 IEEE 17th International Symposium on Biomedical Imaging Workshops (ISBI Workshops), Iowa City, IA, USA, 4 April 2020; pp. 1–4. [Google Scholar]Gulrajani, I.; Ahmed, F.; Arjovsky, M.; Dumoulin, V.; Courville, A. Improved training of Wasserstein GANs. Adv. Neural Inf. Process. Syst. 2017, 30, 5767–5777. [Google Scholar] Diagnostics 13 01306 g001 550

Figure 1. The fundamental structure of the convolutional neural network (CNN): (a) traditional CNN and (b) residual CNN. Figure 1. The fundamental structure of the convolutional neural network (CNN): (a) traditional CNN and (b) residual CNN. Diagnostics 13 01306 g001

Figure 2. Flowchart of the proposed dual-domain magnetic resonance imaging (MRI) reconstruction technique. Figure 2. Flowchart of the proposed dual-domain magnetic resonance imaging (MRI) reconstruction technique. Diagnostics 13 01306 g002

Figure 3. Proposed dual-domain residual attention convolutional neural network (RA-CNN) architecture. Figure 3. Proposed dual-domain residual attention convolutional neural network (RA-CNN) architecture. Diagnostics 13 01306 g003

Figure 4. The attention gate schematic diagram. Figure 4. The attention gate schematic diagram. Diagnostics 13 01306 g004

Figure 5. Data augmentation: (a) augmented k-space and (b) the corresponding reconstructed image. Figure 5. Data augmentation: (a) augmented k-space and (b) the corresponding reconstructed image. Diagnostics 13 01306 g005

Figure 6. Training and validation losses of the proposed technique. Figure 6. Training and validation losses of the proposed technique. Diagnostics 13 01306 g006

Figure 7. MRI images from the test dataset (slice No. 100) for acceleration factor (AF) 4: (a) fully sampled reference image and (b) undersampled k-space. Image reconstructed by the (c) inverse fast Fourier transform (IFFT), (d) Unet, (e) W-net, and (f) RA-CNN methods. Figure 7. MRI images from the test dataset (slice No. 100) for acceleration factor (AF) 4: (a) fully sampled reference image and (b) undersampled k-space. Image reconstructed by the (c) inverse fast Fourier transform (IFFT), (d) Unet, (e) W-net, and (f) RA-CNN methods. Diagnostics 13 01306 g007

Figure 8. MRI images from the test dataset (slice No. 100) for AF 5: (a) fully sampled reference image and (b) undersampled k-space. Image reconstructed by the (c) IFFT, (d) Unet, (e) W-net, and (f) RA-CNN methods. Figure 8. MRI images from the test dataset (slice No. 100) for AF 5: (a) fully sampled reference image and (b) undersampled k-space. Image reconstructed by the (c) IFFT, (d) Unet, (e) W-net, and (f) RA-CNN methods. Diagnostics 13 01306 g008

Figure 9. MRI images from the test dataset (slice No. 100) for AF 6: (a) fully sampled reference image and (b) undersampled k-space. Image reconstructed by the (c) IFFT, (d) Unet, (e) W-net, and (f) RA-CNN methods. Figure 9. MRI images from the test dataset (slice No. 100) for AF 6: (a) fully sampled reference image and (b) undersampled k-space. Image reconstructed by the (c) IFFT, (d) Unet, (e) W-net, and (f) RA-CNN methods. Diagnostics 13 01306 g009

Table 1. Mean ± standard deviation SSIM, NRMSE, and PSNR values obtained using the state-of-the-art methods with different AFs. Table 1. Mean ± standard deviation SSIM, NRMSE, and PSNR values obtained using the state-of-the-art methods with different AFs. AFsMethodsTypeSSIMNRMSEPSNR Zero-filling 0.654 ± 0.040.038 ± 0.0823.93 ± 2.62 dAUTOMAPDM0.849 ± 0.040.025 ± 0.0227.87 ± 1.93 UnetSD0.977 ± 0.060.023 ± 0.0133.28 ± 3.14 DAGANSD0.963 ± 0.100.029 ± 0.0231.33 ± 3.11 RefineGANSD0.979 ± 0.070.019 ± 0.0135.44 ± 3.71 PBCUSD0.983 ± 0.060.018 ± 0.0334.69 ± 3.974×FDA-CNNSD0.981 ± 0.050.012 ± 0.0138.86 ± 4.05 DC-CNNDD0.986 ± 0.050.012 ± 0.0139.51 ± 3.35 KIKI-netDD0.986 ± 0.060.012 ± 0.0139.64 ± 3.35 W-netDD0.985 ± 0.060.014 ± 0.0138.23 ± 3.32 Hybrid cascadeDD0.986 ± 0.030.012 ± 0.0139.87 ± 3.38 Dual-encoder UnetDD0.980 ± 0.050.018 ± 0.0234.53 ± 1.35 RA-CNNDD0.989 ± 0.030.013 ± 0.0041.95 ± 3.21 Zero-filling 0.593 ± 0.050.055 ± 0.0123.63 ± 2.67 dAUTOMAPDM0.823 ± 0.020.051 ± 0.2527.20 ± 1.51 UnetSD0.966 ± 0.100.027 ± 0.0231.88 ± 3.13 DAGANSD0.949 ± 0.110.039 ± 0.0128.69 ± 2.66 RefineGANSD0.973 ± 0.080.023 ± 0.0133.84 ± 3.83 PBCUSD0.983 ± 0.060.021 ± 0.0233.25 ± 2.765×FDA-CNNSD0.976 ± 0.050.016 ± 0.0133.31 ± 4.06 DC-CNNDD0.982 ± 0.070.015 ± 0.0137.67 ± 3.20 KIKI-netDD0.982 ± 0.070.015 ± 0.0237.67 ± 3.22 W-netDD0.981 ± 0.060.017 ± 0.0136.50 ± 3.21 Hybrid cascade DD0.982 ± 0.080.014 ± 0.0237.88 ± 3.25 Dual-encoder Unet DD0.975 ± 0.040.025 ± 0.0333.24 ± 1.70 RA-CNNDD0.986 ± 0.040.015 ± 0.0141.11 ± 3.23 DM: direct mapping; SD: single-domain network; DD: dual-domain network. dAUTOMAP: decomposing automated transform by manifold approximation; DAGAN: de-aliasing generative adversarial network; PBCU: projection-based cascade Unet; CNN: convolutional neural network; FDA-CNN: fully dense attention CNN; DC-CNN: deep cascade CNN; RA-CNN: residual attention CNN; AF: acceleration factor; SSIM: structural similarity index measure; NRMSE: normalized root mean square error; PSNR: peak signal-to-noise ratio. Table

Table 2. Mean ± standard deviation SSIM, NRMSE, and PSNR values of the state-of-the-art methods obtained using AF 6. Table 2. Mean ± standard deviation SSIM, NRMSE, and PSNR values of the state-of-the-art methods obtained using AF 6. MethodsSSIMNRMSEPSNRParameters (M)Zero-filling0.651 ± 0.090.071 ± 0.0820.63 ± 3.15-dAutomap0.692 ± 0.090.033 ± 0.0324.73 ± 3.010.16Unet0.701 ± 0.090.031 ± 0.0227.43 ± 2.493.13PBCU0.801 ± 0.050.035 ± 0.0731.25 ± 1.433.15FDA-CNN0.816 ± 0.180.022 ± 0.0231.88 ± 4.001.01DC-CNN0.779 ± 0.120.030 ± 0.0330.31 ± 3.401.39KIKI-net0.756 ± 0.050.026 ± 0.0930.93 ± 1.491.25W-net0.731 ± 0.070.038 ± 0.0230.70 ± 3.661.13Hybrid cascade0.813 ± 0.110.025 ± 0.1231.19 ± 2.401.66Dual-encoder Unet0.751 ± 0.080.024 ± 0.0225.68 ± 3.000.99RA-CNN0.848 ± 0.190.021 ± 0.0437.30 ± 4.040.68 M = million. Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Share and Cite MDPI and ACS Style

Hossain, M.B.; Kwon, K.-C.; Shinde, R.K.; Imtiaz, S.M.; Kim, N. A Hybrid Residual Attention Convolutional Neural Network for Compressed Sensing Magnetic Resonance Image Reconstruction. Diagnostics 2023, 13, 1306. https://doi.org/10.3390/diagnostics13071306

AMA Style

Hossain MB, Kwon K-C, Shinde RK, Imtiaz SM, Kim N. A Hybrid Residual Attention Convolutional Neural Network for Compressed Sensing Magnetic Resonance Image Reconstruction. Diagnostics. 2023; 13(7):1306. https://doi.org/10.3390/diagnostics13071306

Chicago/Turabian Style

Hossain, Md. Biddut, Ki-Chul Kwon, Rupali Kiran Shinde, Shariar Md Imtiaz, and Nam Kim. 2023. "A Hybrid Residual Attention Convolutional Neural Network for Compressed Sensing Magnetic Resonance Image Reconstruction" Diagnostics 13, no. 7: 1306. https://doi.org/10.3390/diagnostics13071306

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here. Article Metrics No No Article Access Statistics For more information on the journal statistics, click here. Multiple requests from the same IP address are counted as one view. Zoom | Orient | As Lines | As Sticks | As Cartoon | As Surface | Previous Scene | Next Scene Cite Export citation file: BibTeX | EndNote | RIS MDPI and ACS Style

AMA Style

Chicago/Turabian Style

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here. clear Diagnostics, EISSN 2075-4418, Published by MDPI RSS Content Alert Further Information Article Processing Charges Pay an Invoice Open Access Policy Contact MDPI Jobs at MDPI Guidelines For Authors For Reviewers For Editors For Librarians For Publishers For Societies For Conference Organizers MDPI Initiatives Sciforum MDPI Books Preprints.org Scilit SciProfiles Encyclopedia JAMS Proceedings Series Follow MDPI LinkedIn Facebook Twitter

© 1996-2023 MDPI (Basel, Switzerland) unless otherwise stated Disclaimer Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. Terms and Conditions Privacy Policy We use cookies on our website to ensure you get the best experience. Read more about our cookies here. Accept Share Link Copy clear Share https://www.mdpi.com/2224312 clear Back to TopTop

【本文地址】

Diagnostics

Diagnostics

今日新闻

推荐新闻