File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Development of fast deep learning quantification for magnetic resonance fingerprinting in vivo

TitleDevelopment of fast deep learning quantification for magnetic resonance fingerprinting in vivo
Authors
KeywordsDeep learning
Relaxometry
Brain
Liver
Prostate
Issue Date2020
PublisherElsevier Inc. The Journal's web site is located at http://www.elsevier.com/locate/mri
Citation
Magnetic Resonance Imaging, 2020, v. 70, p. 81-90 How to Cite?
AbstractPurpose: A deep neural network was developed for magnetic resonance fingerprinting (MRF) quantification. This study aimed at extending previous studies of deep learning MRF to in vivo applications, allowing sub-second computation time for large-scale data. Methods: We applied the deep learning methodology based on our previously published multi-layer perceptron. The number of layers was four, which was optimized to balance the model capacity and noise robustness. The training sets were obtained from MRF dictionaries with 9000 to 28,000 atoms, depending on the desired T1 and T2 ranges. The simulated MRF undersampling artifact based on the k-space acquisition scheme and noise were both added to the training data to reduce the error in estimates. Results: The neural network achieved high fidelity (R2 _ 0.98) as compared to the T1 and T2 values of the ISMRM standardized phantom. In brain MRF experiment, the model trained with simulated artifacts and noise showed less error compared to that without. The in vivo application of our neural network for liver and prostate were also demonstrated. For an MRF slice with 256 _ 256 image resolution, the computation time of our neural network was 0.12 s, compared with the _ 28 s-pre-slice for the conventional dictionary matching method. Conclusion: Our neural network achieved fast computation speed for MRF quantification. The model trained with simulated artifacts and noise showed less error and achieved optimal performance for phantom experiment and in vivo normal brain and liver, and prostate cancer patient.
Persistent Identifierhttp://hdl.handle.net/10722/283046
ISSN
2019 Impact Factor: 2.053
2015 SCImago Journal Rankings: 1.139

 

DC FieldValueLanguage
dc.contributor.authorCao, P-
dc.contributor.authorCui, D-
dc.contributor.authorVardhanabhuti, V-
dc.contributor.authorHui, ES-
dc.date.accessioned2020-06-05T06:24:21Z-
dc.date.available2020-06-05T06:24:21Z-
dc.date.issued2020-
dc.identifier.citationMagnetic Resonance Imaging, 2020, v. 70, p. 81-90-
dc.identifier.issn0730-725X-
dc.identifier.urihttp://hdl.handle.net/10722/283046-
dc.description.abstractPurpose: A deep neural network was developed for magnetic resonance fingerprinting (MRF) quantification. This study aimed at extending previous studies of deep learning MRF to in vivo applications, allowing sub-second computation time for large-scale data. Methods: We applied the deep learning methodology based on our previously published multi-layer perceptron. The number of layers was four, which was optimized to balance the model capacity and noise robustness. The training sets were obtained from MRF dictionaries with 9000 to 28,000 atoms, depending on the desired T1 and T2 ranges. The simulated MRF undersampling artifact based on the k-space acquisition scheme and noise were both added to the training data to reduce the error in estimates. Results: The neural network achieved high fidelity (R2 _ 0.98) as compared to the T1 and T2 values of the ISMRM standardized phantom. In brain MRF experiment, the model trained with simulated artifacts and noise showed less error compared to that without. The in vivo application of our neural network for liver and prostate were also demonstrated. For an MRF slice with 256 _ 256 image resolution, the computation time of our neural network was 0.12 s, compared with the _ 28 s-pre-slice for the conventional dictionary matching method. Conclusion: Our neural network achieved fast computation speed for MRF quantification. The model trained with simulated artifacts and noise showed less error and achieved optimal performance for phantom experiment and in vivo normal brain and liver, and prostate cancer patient.-
dc.languageeng-
dc.publisherElsevier Inc. The Journal's web site is located at http://www.elsevier.com/locate/mri-
dc.relation.ispartofMagnetic Resonance Imaging-
dc.subjectDeep learning-
dc.subjectRelaxometry-
dc.subjectBrain-
dc.subjectLiver-
dc.subjectProstate-
dc.titleDevelopment of fast deep learning quantification for magnetic resonance fingerprinting in vivo-
dc.typeArticle-
dc.identifier.emailCao, P: caopeng1@hku.hk-
dc.identifier.emailCui, D: cuidi00@hku.hk-
dc.identifier.emailVardhanabhuti, V: varv@hku.hk-
dc.identifier.emailHui, ES: edshui@hku.hk-
dc.identifier.authorityCao, P=rp02474-
dc.identifier.authorityVardhanabhuti, V=rp01900-
dc.identifier.authorityHui, ES=rp01832-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1016/j.mri.2020.03.009-
dc.identifier.pmid32276007-
dc.identifier.scopuseid_2-s2.0-85084069620-
dc.identifier.hkuros310276-
dc.identifier.volume70-
dc.identifier.spage81-
dc.identifier.epage90-
dc.publisher.placeUnited States-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats