Acceleration of Magnetic Resonance Fingerprinting Reconstruction Using Denoising and Self-Attention Pyramidal Convolutional Neural Network
Magnetic resonance fingerprinting (MRF) based on echo-planar imaging (EPI) enables whole-brain imaging to rapidly obtain T1 and T2* relaxation time maps. Reconstructing parametric maps from the MRF scanned baselines by the inner-product method is computationally expensive. We aimed to accelerate the...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-02-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/22/3/1260 |
| _version_ | 1827658627336372224 |
|---|---|
| author | Jia-Sheng Hong Ingo Hermann Frank Gerrit Zöllner Lothar R. Schad Shuu-Jiun Wang Wei-Kai Lee Yung-Lin Chen Yu Chang Yu-Te Wu |
| author_facet | Jia-Sheng Hong Ingo Hermann Frank Gerrit Zöllner Lothar R. Schad Shuu-Jiun Wang Wei-Kai Lee Yung-Lin Chen Yu Chang Yu-Te Wu |
| author_sort | Jia-Sheng Hong |
| collection | DOAJ |
| description | Magnetic resonance fingerprinting (MRF) based on echo-planar imaging (EPI) enables whole-brain imaging to rapidly obtain T1 and T2* relaxation time maps. Reconstructing parametric maps from the MRF scanned baselines by the inner-product method is computationally expensive. We aimed to accelerate the reconstruction of parametric maps for MRF-EPI by using a deep learning model. The proposed approach uses a two-stage model that first eliminates noise and then regresses the parametric maps. Parametric maps obtained by dictionary matching were used as a reference and compared with the prediction results of the two-stage model. MRF-EPI scans were collected from 32 subjects. The signal-to-noise ratio increased significantly after the noise removal by the denoising model. For prediction with scans in the testing dataset, the mean absolute percentage errors between the standard and the final two-stage model were 3.1%, 3.2%, and 1.9% for T1, and 2.6%, 2.3%, and 2.8% for T2* in gray matter, white matter, and lesion locations, respectively. Our proposed two-stage deep learning model can effectively remove noise and accurately reconstruct MRF-EPI parametric maps, increasing the speed of reconstruction and reducing the storage space required by dictionaries. |
| first_indexed | 2024-03-09T23:06:18Z |
| format | Article |
| id | doaj.art-db277e97b54e49ec8def22b5e3cb9322 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T23:06:18Z |
| publishDate | 2022-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-db277e97b54e49ec8def22b5e3cb93222023-11-23T17:53:07ZengMDPI AGSensors1424-82202022-02-01223126010.3390/s22031260Acceleration of Magnetic Resonance Fingerprinting Reconstruction Using Denoising and Self-Attention Pyramidal Convolutional Neural NetworkJia-Sheng Hong0Ingo Hermann1Frank Gerrit Zöllner2Lothar R. Schad3Shuu-Jiun Wang4Wei-Kai Lee5Yung-Lin Chen6Yu Chang7Yu-Te Wu8Department of Biomedical Imaging and Radiological Sciences, National Yang Ming Chiao Tung University, Taipei 112, TaiwanComputer Assisted Clinical Medicine, Mannheim Institute for Intelligent Systems in Medicine, Medical Faculty Mannheim, Heidelberg University, 68167 Mannheim, GermanyComputer Assisted Clinical Medicine, Mannheim Institute for Intelligent Systems in Medicine, Medical Faculty Mannheim, Heidelberg University, 68167 Mannheim, GermanyComputer Assisted Clinical Medicine, Mannheim Institute for Intelligent Systems in Medicine, Medical Faculty Mannheim, Heidelberg University, 68167 Mannheim, GermanyDepartment of Neurology, Neurological Institute, Taipei Veterans General Hospital, Taipei 112, TaiwanDepartment of Biomedical Imaging and Radiological Sciences, National Yang Ming Chiao Tung University, Taipei 112, TaiwanInstitute of Biophotonics, National Yang Ming Chiao Tung University, Taipei 112, TaiwanInstitute of Biophotonics, National Yang Ming Chiao Tung University, Taipei 112, TaiwanBrain Research Center, National Yang Ming Chiao Tung University, Taipei 112, TaiwanMagnetic resonance fingerprinting (MRF) based on echo-planar imaging (EPI) enables whole-brain imaging to rapidly obtain T1 and T2* relaxation time maps. Reconstructing parametric maps from the MRF scanned baselines by the inner-product method is computationally expensive. We aimed to accelerate the reconstruction of parametric maps for MRF-EPI by using a deep learning model. The proposed approach uses a two-stage model that first eliminates noise and then regresses the parametric maps. Parametric maps obtained by dictionary matching were used as a reference and compared with the prediction results of the two-stage model. MRF-EPI scans were collected from 32 subjects. The signal-to-noise ratio increased significantly after the noise removal by the denoising model. For prediction with scans in the testing dataset, the mean absolute percentage errors between the standard and the final two-stage model were 3.1%, 3.2%, and 1.9% for T1, and 2.6%, 2.3%, and 2.8% for T2* in gray matter, white matter, and lesion locations, respectively. Our proposed two-stage deep learning model can effectively remove noise and accurately reconstruct MRF-EPI parametric maps, increasing the speed of reconstruction and reducing the storage space required by dictionaries.https://www.mdpi.com/1424-8220/22/3/1260magnetic resonance fingerprintingecho-planar imagingT1 and T2* relaxation timesdenoising convolutional neural networkself-attentionfeature pyramid network |
| spellingShingle | Jia-Sheng Hong Ingo Hermann Frank Gerrit Zöllner Lothar R. Schad Shuu-Jiun Wang Wei-Kai Lee Yung-Lin Chen Yu Chang Yu-Te Wu Acceleration of Magnetic Resonance Fingerprinting Reconstruction Using Denoising and Self-Attention Pyramidal Convolutional Neural Network Sensors magnetic resonance fingerprinting echo-planar imaging T1 and T2* relaxation times denoising convolutional neural network self-attention feature pyramid network |
| title | Acceleration of Magnetic Resonance Fingerprinting Reconstruction Using Denoising and Self-Attention Pyramidal Convolutional Neural Network |
| title_full | Acceleration of Magnetic Resonance Fingerprinting Reconstruction Using Denoising and Self-Attention Pyramidal Convolutional Neural Network |
| title_fullStr | Acceleration of Magnetic Resonance Fingerprinting Reconstruction Using Denoising and Self-Attention Pyramidal Convolutional Neural Network |
| title_full_unstemmed | Acceleration of Magnetic Resonance Fingerprinting Reconstruction Using Denoising and Self-Attention Pyramidal Convolutional Neural Network |
| title_short | Acceleration of Magnetic Resonance Fingerprinting Reconstruction Using Denoising and Self-Attention Pyramidal Convolutional Neural Network |
| title_sort | acceleration of magnetic resonance fingerprinting reconstruction using denoising and self attention pyramidal convolutional neural network |
| topic | magnetic resonance fingerprinting echo-planar imaging T1 and T2* relaxation times denoising convolutional neural network self-attention feature pyramid network |
| url | https://www.mdpi.com/1424-8220/22/3/1260 |
| work_keys_str_mv | AT jiashenghong accelerationofmagneticresonancefingerprintingreconstructionusingdenoisingandselfattentionpyramidalconvolutionalneuralnetwork AT ingohermann accelerationofmagneticresonancefingerprintingreconstructionusingdenoisingandselfattentionpyramidalconvolutionalneuralnetwork AT frankgerritzollner accelerationofmagneticresonancefingerprintingreconstructionusingdenoisingandselfattentionpyramidalconvolutionalneuralnetwork AT lotharrschad accelerationofmagneticresonancefingerprintingreconstructionusingdenoisingandselfattentionpyramidalconvolutionalneuralnetwork AT shuujiunwang accelerationofmagneticresonancefingerprintingreconstructionusingdenoisingandselfattentionpyramidalconvolutionalneuralnetwork AT weikailee accelerationofmagneticresonancefingerprintingreconstructionusingdenoisingandselfattentionpyramidalconvolutionalneuralnetwork AT yunglinchen accelerationofmagneticresonancefingerprintingreconstructionusingdenoisingandselfattentionpyramidalconvolutionalneuralnetwork AT yuchang accelerationofmagneticresonancefingerprintingreconstructionusingdenoisingandselfattentionpyramidalconvolutionalneuralnetwork AT yutewu accelerationofmagneticresonancefingerprintingreconstructionusingdenoisingandselfattentionpyramidalconvolutionalneuralnetwork |