1
|
Huang J, Yang L, Wang F, Wu Y, Nan Y, Wu W, Wang C, Shi K, Aviles-Rivero AI, Schönlieb CB, Zhang D, Yang G. Enhancing global sensitivity and uncertainty quantification in medical image reconstruction with Monte Carlo arbitrary-masked mamba. Med Image Anal 2024; 99:103334. [PMID: 39255733 DOI: 10.1016/j.media.2024.103334] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/18/2024] [Revised: 08/05/2024] [Accepted: 09/01/2024] [Indexed: 09/12/2024]
Abstract
Deep learning has been extensively applied in medical image reconstruction, where Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs) represent the predominant paradigms, each possessing distinct advantages and inherent limitations: CNNs exhibit linear complexity with local sensitivity, whereas ViTs demonstrate quadratic complexity with global sensitivity. The emerging Mamba has shown superiority in learning visual representation, which combines the advantages of linear scalability and global sensitivity. In this study, we introduce MambaMIR, an Arbitrary-Masked Mamba-based model with wavelet decomposition for joint medical image reconstruction and uncertainty estimation. A novel Arbitrary Scan Masking (ASM) mechanism "masks out" redundant information to introduce randomness for further uncertainty estimation. Compared to the commonly used Monte Carlo (MC) dropout, our proposed MC-ASM provides an uncertainty map without the need for hyperparameter tuning and mitigates the performance drop typically observed when applying dropout to low-level tasks. For further texture preservation and better perceptual quality, we employ the wavelet transformation into MambaMIR and explore its variant based on the Generative Adversarial Network, namely MambaMIR-GAN. Comprehensive experiments have been conducted for multiple representative medical image reconstruction tasks, demonstrating that the proposed MambaMIR and MambaMIR-GAN outperform other baseline and state-of-the-art methods in different reconstruction tasks, where MambaMIR achieves the best reconstruction fidelity and MambaMIR-GAN has the best perceptual quality. In addition, our MC-ASM provides uncertainty maps as an additional tool for clinicians, while mitigating the typical performance drop caused by the commonly used dropout.
Collapse
Affiliation(s)
- Jiahao Huang
- Bioengineering Department and Imperial-X, Imperial College London, London W12 7SL, United Kingdom; National Heart and Lung Institute, Imperial College London, London SW7 2AZ, United Kingdom; Cardiovascular Research Centre, Royal Brompton Hospital, London SW3 6NP, United Kingdom.
| | - Liutao Yang
- College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, China
| | - Fanwen Wang
- Bioengineering Department and Imperial-X, Imperial College London, London W12 7SL, United Kingdom; National Heart and Lung Institute, Imperial College London, London SW7 2AZ, United Kingdom; Cardiovascular Research Centre, Royal Brompton Hospital, London SW3 6NP, United Kingdom
| | - Yinzhe Wu
- Bioengineering Department and Imperial-X, Imperial College London, London W12 7SL, United Kingdom; National Heart and Lung Institute, Imperial College London, London SW7 2AZ, United Kingdom; Cardiovascular Research Centre, Royal Brompton Hospital, London SW3 6NP, United Kingdom.
| | - Yang Nan
- Bioengineering Department and Imperial-X, Imperial College London, London W12 7SL, United Kingdom; National Heart and Lung Institute, Imperial College London, London SW7 2AZ, United Kingdom
| | - Weiwen Wu
- School of Biomedical Engineering, Shenzhen Campus of Sun Yat-sen University, Guangdong, China
| | - Chengyan Wang
- Human Phenome Institute, Fudan University, Shanghai, China
| | - Kuangyu Shi
- Department of Nuclear Medicine, Inselspital, University of Bern, Bern, Switzerland; Department of Informatics, Technical University of Munich, Munich, Germany
| | - Angelica I Aviles-Rivero
- Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Cambridge, United Kingdom
| | - Carola-Bibiane Schönlieb
- Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Cambridge, United Kingdom
| | - Daoqiang Zhang
- College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, China
| | - Guang Yang
- Bioengineering Department and Imperial-X, Imperial College London, London W12 7SL, United Kingdom; National Heart and Lung Institute, Imperial College London, London SW7 2AZ, United Kingdom; Cardiovascular Research Centre, Royal Brompton Hospital, London SW3 6NP, United Kingdom; School of Biomedical Engineering & Imaging Sciences, King's College London, London WC2R 2LS, United Kingdom.
| |
Collapse
|
2
|
Wang S, Wu R, Jia S, Diakite A, Li C, Liu Q, Zheng H, Ying L. Knowledge-driven deep learning for fast MR imaging: Undersampled MR image reconstruction from supervised to un-supervised learning. Magn Reson Med 2024; 92:496-518. [PMID: 38624162 DOI: 10.1002/mrm.30105] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/03/2023] [Revised: 03/19/2024] [Accepted: 03/20/2024] [Indexed: 04/17/2024]
Abstract
Deep learning (DL) has emerged as a leading approach in accelerating MRI. It employs deep neural networks to extract knowledge from available datasets and then applies the trained networks to reconstruct accurate images from limited measurements. Unlike natural image restoration problems, MRI involves physics-based imaging processes, unique data properties, and diverse imaging tasks. This domain knowledge needs to be integrated with data-driven approaches. Our review will introduce the significant challenges faced by such knowledge-driven DL approaches in the context of fast MRI along with several notable solutions, which include learning neural networks and addressing different imaging application scenarios. The traits and trends of these techniques have also been given which have shifted from supervised learning to semi-supervised learning, and finally, to unsupervised learning methods. In addition, MR vendors' choices of DL reconstruction have been provided along with some discussions on open questions and future directions, which are critical for the reliable imaging systems.
Collapse
Affiliation(s)
- Shanshan Wang
- Paul C Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
| | - Ruoyou Wu
- Paul C Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
| | - Sen Jia
- Paul C Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
| | - Alou Diakite
- Paul C Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
- University of Chinese Academy of Sciences, Beijing, China
| | - Cheng Li
- Paul C Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
| | - Qiegen Liu
- Department of Electronic Information Engineering, Nanchang University, Nanchang, China
| | - Hairong Zheng
- Paul C Lauterbur Research Center for Biomedical Imaging, Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China
| | - Leslie Ying
- Department of Biomedical Engineering and Department of Electrical Engineering, The State University of New York, Buffalo, New York, USA
| |
Collapse
|
3
|
Wang M, Ma Y, Li L, Pan X, Wen Y, Qiu Y, Guo D, Zhu Y, Lian J, Tong D. Compressed Sensitivity Encoding Artificial Intelligence Accelerates Brain Metastasis Imaging by Optimizing Image Quality and Reducing Scan Time. AJNR Am J Neuroradiol 2024; 45:444-452. [PMID: 38485196 PMCID: PMC11288577 DOI: 10.3174/ajnr.a8161] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/11/2023] [Accepted: 12/25/2023] [Indexed: 04/10/2024]
Abstract
BACKGROUND AND PURPOSE Accelerating the image acquisition speed of MR imaging without compromising the image quality is challenging. This study aimed to evaluate the feasibility of contrast-enhanced (CE) 3D T1WI and CE 3D-FLAIR sequences reconstructed with compressed sensitivity encoding artificial intelligence (CS-AI) for detecting brain metastases (BM) and explore the optimal acceleration factor (AF) for clinical BM imaging. MATERIALS AND METHODS Fifty-one patients with cancer with suspected BM were included. Fifty participants underwent different customized CE 3D-T1WI or CE 3D-FLAIR sequence scans. Compressed SENSE encoding acceleration 6 (CS6), a commercially available standard sequence, was used as the reference standard. Quantitative and qualitative methods were used to evaluate image quality. The SNR and contrast-to-noise ratio (CNR) were calculated, and qualitative evaluations were independently conducted by 2 neuroradiologists. After exploring the optimal AF, sample images were obtained from 1 patient by using both optimized sequences. RESULTS Quantitatively, the CNR of the CS-AI protocol for CE 3D-T1WI and CE 3D-FLAIR sequences was superior to that of the CS protocol under the same AF (P < .05). Compared with reference CS6, the CS-AI groups had higher CNR values (all P < .05), with the CS-AI10 scan having the highest value. The SNR of the CS-AI group was better than that of the reference for both CE 3D-T1WI and CE 3D-FLAIR sequences (all P < .05). Qualitatively, the CS-AI protocol produced higher image quality scores than did the CS protocol with the same AF (all P < .05). In contrast to the reference CS6, the CS-AI group showed good image quality scores until an AF of up to 10 (all P < .05). The CS-AI10 scan provided the optimal images, improving the delineation of normal gray-white matter boundaries and lesion areas (P < .05). Compared with the reference, CS-AI10 showed reductions in scan time of 39.25% and 39.93% for CE 3D-T1WI and CE 3D-FLAIR sequences, respectively. CONCLUSIONS CE 3D-T1WI and CE 3D-FLAIR sequences reconstructed with CS-AI for the detection of BM may provide a more effective alternative reconstruction approach than CS. CS-AI10 is suitable for clinical applications, providing optimal image quality and a shortened scan time.
Collapse
Affiliation(s)
- Mengmeng Wang
- From the Department of Radiology (M.W., Y.M., L.L., X.P., Y.W., Y.Q., D.G., D.T.), The First Hospital of Jilin University, Changchun, China
| | - Yue Ma
- From the Department of Radiology (M.W., Y.M., L.L., X.P., Y.W., Y.Q., D.G., D.T.), The First Hospital of Jilin University, Changchun, China
| | - Linna Li
- From the Department of Radiology (M.W., Y.M., L.L., X.P., Y.W., Y.Q., D.G., D.T.), The First Hospital of Jilin University, Changchun, China
| | - Xingchen Pan
- From the Department of Radiology (M.W., Y.M., L.L., X.P., Y.W., Y.Q., D.G., D.T.), The First Hospital of Jilin University, Changchun, China
| | - Yafei Wen
- From the Department of Radiology (M.W., Y.M., L.L., X.P., Y.W., Y.Q., D.G., D.T.), The First Hospital of Jilin University, Changchun, China
| | - Ying Qiu
- From the Department of Radiology (M.W., Y.M., L.L., X.P., Y.W., Y.Q., D.G., D.T.), The First Hospital of Jilin University, Changchun, China
| | - Dandan Guo
- From the Department of Radiology (M.W., Y.M., L.L., X.P., Y.W., Y.Q., D.G., D.T.), The First Hospital of Jilin University, Changchun, China
| | - Yi Zhu
- Philips Healthcare (Y.Z., J.L., D.T.), Beijing, China
| | - Jianxiu Lian
- Philips Healthcare (Y.Z., J.L., D.T.), Beijing, China
| | - Dan Tong
- From the Department of Radiology (M.W., Y.M., L.L., X.P., Y.W., Y.Q., D.G., D.T.), The First Hospital of Jilin University, Changchun, China
| |
Collapse
|
4
|
Li Z, Wang Y, Zhang J, Wu W, Yu H. Two-and-a-half order score-based model for solving 3D ill-posed inverse problems. Comput Biol Med 2024; 168:107819. [PMID: 38064853 DOI: 10.1016/j.compbiomed.2023.107819] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/28/2023] [Revised: 11/25/2023] [Accepted: 12/03/2023] [Indexed: 01/10/2024]
Abstract
Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) are crucial technologies in the field of medical imaging. Score-based models demonstrated effectiveness in addressing different inverse problems encountered in the field of CT and MRI, such as sparse-view CT and fast MRI reconstruction. However, these models face challenges in achieving accurate three dimensional (3D) volumetric reconstruction. The existing score-based models predominantly concentrate on reconstructing two-dimensional (2D) data distributions, resulting in inconsistencies between adjacent slices in the reconstructed 3D volumetric images. To overcome this limitation, we propose a novel two-and-a-half order score-based model (TOSM). During the training phase, our TOSM learns data distributions in 2D space, simplifying the training process compared to working directly on 3D volumes. However, during the reconstruction phase, the TOSM utilizes complementary scores along three directions (sagittal, coronal, and transaxial) to achieve a more precise reconstruction. The development of TOSM is built on robust theoretical principles, ensuring its reliability and efficacy. Through extensive experimentation on large-scale sparse-view CT and fast MRI datasets, our method achieved state-of-the-art (SOTA) results in solving 3D ill-posed inverse problems, averaging a 1.56 dB peak signal-to-noise ratio (PSNR) improvement over existing sparse-view CT reconstruction methods across 29 views and 0.87 dB PSNR improvement over existing fast MRI reconstruction methods with × 2 acceleration. In summary, TOSM significantly addresses the issue of inconsistency in 3D ill-posed problems by modeling the distribution of 3D data rather than 2D distribution which has achieved remarkable results in both CT and MRI reconstruction tasks.
Collapse
Affiliation(s)
- Zirong Li
- Department of Biomedical Engineering, Sun-Yat-sen University, Shenzhen Campus, Shenzhen, China
| | - Yanyang Wang
- Department of Biomedical Engineering, Sun-Yat-sen University, Shenzhen Campus, Shenzhen, China
| | - Jianjia Zhang
- Department of Biomedical Engineering, Sun-Yat-sen University, Shenzhen Campus, Shenzhen, China
| | - Weiwen Wu
- Department of Biomedical Engineering, Sun-Yat-sen University, Shenzhen Campus, Shenzhen, China.
| | - Hengyong Yu
- Department of Electrical and Computer Engineering, University of Massachusetts Lowell, Lowell, MA, USA.
| |
Collapse
|
5
|
Wang S, Luo Y, Zuo P, Pan L, Li Y, Sun Z. In-memory analog solution of compressed sensing recovery in one step. SCIENCE ADVANCES 2023; 9:eadj2908. [PMID: 38091396 PMCID: PMC10848716 DOI: 10.1126/sciadv.adj2908] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 06/17/2023] [Accepted: 11/13/2023] [Indexed: 02/12/2024]
Abstract
Modern analog computing, by gaining momentum from nonvolatile resistive memory devices, deals with matrix computations. In-memory analog computing has been demonstrated for solving some basic but ordinary matrix problems in one step. Among the more complicated matrix problems, compressed sensing (CS) is a prominent example, whose recovery algorithms feature high-order matrix operations and hardware-unfriendly nonlinear functions. In light of the local competitive algorithm (LCA), here, we present a closed-loop, continuous-time resistive memory circuit for solving CS recovery in one step. Recovery of one-dimensional (1D) sparse signal and 2D compressive images has been experimentally demonstrated, showing elapsed times around few microseconds and normalized mean squared errors of 10-2. The LCA circuit is one or two orders of magnitude faster than conventional digital approaches. It also substantially outperforms other (electronic or exotically photonic) analog CS recovery methods in terms of speed, energy, and fidelity, thus representing a highly promising technology for real-time CS applications.
Collapse
Affiliation(s)
- Shiqing Wang
- School of Integrated Circuits, Institute for Artificial Intelligence and Beijing Advanced Innovation Center for Integrated Circuits, Peking University, Beijing 100871, China
| | - Yubiao Luo
- School of Integrated Circuits, Institute for Artificial Intelligence and Beijing Advanced Innovation Center for Integrated Circuits, Peking University, Beijing 100871, China
| | - Pushen Zuo
- School of Integrated Circuits, Institute for Artificial Intelligence and Beijing Advanced Innovation Center for Integrated Circuits, Peking University, Beijing 100871, China
| | - Lunshuai Pan
- School of Integrated Circuits, Institute for Artificial Intelligence and Beijing Advanced Innovation Center for Integrated Circuits, Peking University, Beijing 100871, China
| | - Yongxiang Li
- School of Integrated Circuits, Institute for Artificial Intelligence and Beijing Advanced Innovation Center for Integrated Circuits, Peking University, Beijing 100871, China
| | - Zhong Sun
- School of Integrated Circuits, Institute for Artificial Intelligence and Beijing Advanced Innovation Center for Integrated Circuits, Peking University, Beijing 100871, China
| |
Collapse
|
6
|
Fujita N, Yokosawa S, Shirai T, Terada Y. Numerical and Clinical Evaluation of the Robustness of Open-source Networks for Parallel MR Imaging Reconstruction. Magn Reson Med Sci 2023:mp.2023-0031. [PMID: 37518672 DOI: 10.2463/mrms.mp.2023-0031] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 08/01/2023] Open
Abstract
PURPOSE Deep neural networks (DNNs) for MRI reconstruction often require large datasets for training. Still, in clinical settings, the domains of datasets are diverse, and how robust DNNs are to domain differences between training and testing datasets has been an open question. Here, we numerically and clinically evaluate the generalization of the reconstruction networks across various domains under clinically practical conditions and provide practical guidance on what points to consider when selecting models for clinical application. METHODS We compare the reconstruction performance between four network models: U-Net, the deep cascade of convolutional neural networks (DC-CNNs), Hybrid Cascade, and variational network (VarNet). We used the public multicoil dataset fastMRI for training and testing and performed a single-domain test, where the domains of the dataset used for training and testing were the same, and cross-domain tests, where the source and target domains were different. We conducted a single-domain test (Experiment 1) and cross-domain tests (Experiments 2-4), focusing on six factors (the number of images, sampling pattern, acceleration factor, noise level, contrast, and anatomical structure) both numerically and clinically. RESULTS U-Net had lower performance than the three model-based networks and was less robust to domain shifts between training and testing datasets. VarNet had the highest performance and robustness among the three model-based networks, followed by Hybrid Cascade and DC-CNN. Especially, VarNet showed high performance even with a limited number of training images (200 images/10 cases). U-Net was more robust to domain shifts concerning noise level than the other model-based networks. Hybrid Cascade showed slightly better performance and robustness than DC-CNN, except for robustness to noise-level domain shifts. The results of the clinical evaluations generally agreed with the results of the quantitative metrics. CONCLUSION In this study, we numerically and clinically evaluated the robustness of the publicly available networks using the multicoil data. Therefore, this study provided practical guidance for clinical applications.
Collapse
Affiliation(s)
- Naoto Fujita
- Institute of Applied Physics, University of Tsukuba
| | - Suguru Yokosawa
- FUJIFILM Corporation, Medical Systems Research & Development Center
| | - Toru Shirai
- FUJIFILM Corporation, Medical Systems Research & Development Center
| | | |
Collapse
|
7
|
Zhao W, Gao L, Zhai A, Wang D. Comparison of Common Algorithms for Single-Pixel Imaging via Compressed Sensing. SENSORS (BASEL, SWITZERLAND) 2023; 23:4678. [PMID: 37430593 DOI: 10.3390/s23104678] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 03/14/2023] [Revised: 04/19/2023] [Accepted: 05/09/2023] [Indexed: 07/12/2023]
Abstract
Single-pixel imaging (SPI) uses a single-pixel detector instead of a detector array with a lot of pixels in traditional imaging techniques to realize two-dimensional or even multi-dimensional imaging. For SPI using compressed sensing, the target to be imaged is illuminated by a series of patterns with spatial resolution, and then the reflected or transmitted intensity is compressively sampled by the single-pixel detector to reconstruct the target image while breaking the limitation of the Nyquist sampling theorem. Recently, in the area of signal processing using compressed sensing, many measurement matrices as well as reconstruction algorithms have been proposed. It is necessary to explore the application of these methods in SPI. Therefore, this paper reviews the concept of compressive sensing SPI and summarizes the main measurement matrices and reconstruction algorithms in compressive sensing. Further, the performance of their applications in SPI through simulations and experiments is explored in detail, and then their advantages and disadvantages are summarized. Finally, the prospect of compressive sensing with SPI is discussed.
Collapse
Affiliation(s)
- Wenjing Zhao
- College of Physics and Optoelectronics, Taiyuan University of Technology, No. 79 West Main Street, Taiyuan 030024, China
| | - Lei Gao
- College of Physics and Optoelectronics, Taiyuan University of Technology, No. 79 West Main Street, Taiyuan 030024, China
| | - Aiping Zhai
- College of Physics and Optoelectronics, Taiyuan University of Technology, No. 79 West Main Street, Taiyuan 030024, China
| | - Dong Wang
- College of Physics and Optoelectronics, Taiyuan University of Technology, No. 79 West Main Street, Taiyuan 030024, China
- Key Laboratory of Advanced Transducers and Intelligent Control System, Ministry of Education, and Shanxi Province, Taiyuan University of Technology, No. 79 West Main Street, Taiyuan 030024, China
| |
Collapse
|
8
|
Wang G, Fessler JA. Efficient Approximation of Jacobian Matrices Involving a Non-Uniform Fast Fourier Transform (NUFFT). IEEE TRANSACTIONS ON COMPUTATIONAL IMAGING 2023; 9:43-54. [PMID: 37090025 PMCID: PMC10118239 DOI: 10.1109/tci.2023.3240081] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 05/03/2023]
Abstract
There is growing interest in learning Fourier domain sampling strategies (particularly for magnetic resonance imaging, MRI) using optimization approaches. For non-Cartesian sampling, the system models typically involve non-uniform fast Fourier transform (NUFFT) operations. Commonly used NUFFT algorithms contain frequency domain interpolation, which is not differentiable with respect to the sampling pattern, complicating the use of gradient methods. This paper describes an efficient and accurate approach for computing approximate gradients involving NUFFTs. Multiple numerical experiments validate the improved accuracy and efficiency of the proposed approximation. As an application to computational imaging, the NUFFT Jacobians were used to optimize non-Cartesian MRI sampling trajectories via data-driven stochastic optimization. Specifically, the sampling patterns were learned with respect to various model-based image reconstruction (MBIR) algorithms. The proposed approach enables sampling optimization for image sizes that are infeasible with standard auto-differentiation methods due to memory limits. The synergistic acquisition and reconstruction design leads to remarkably improved image quality. In fact, we show that model-based image reconstruction methods with suitably optimized imaging parameters can perform nearly as well as CNN-based methods.
Collapse
Affiliation(s)
- Guanhua Wang
- Department of Biomedical Engineering, University of Michigan, Ann Arbor, MI 48109 USA
| | | |
Collapse
|