Semisupervised Tripled Dictionary Learning for Standard-Dose PET Image Prediction Using Low-Dose PET and Multimodal MRI

Yan Wang, Guangkai Ma, Le An, Feng Shi, Pei Zhang, David S. Lalush, Xi Wu, Yifei Pu, Jiliu Zhou, Dinggang Shen

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

Objective: To obtain high-quality positron emission tomography (PET) image with low-dose tracer injection, this study attempts to predict the standard-dose PET (S-PET) image from both its low-dose PET (L-PET) counterpart and corresponding magnetic resonance imaging (MRI). Methods: It was achieved by patch-based sparse representation (SR), using the training samples with a complete set of MRI, L-PET and S-PET modalities for dictionary construction. However, the number of training samples with complete modalities is often limited. In practice, many samples generally have incomplete modalities (i.e., with one or two missing modalities) that thus cannot be used in the prediction process. In light of this, we develop a semisupervised tripled dictionary learning (SSTDL) method for S-PET image prediction, which can utilize not only the samples with complete modalities (called complete samples) but also the samples with incomplete modalities (called incomplete samples), to take advantage of the large number of available training samples and thus further improve the prediction performance. Results: Validation was done on a real human brain dataset consisting of 18 subjects, and the results show that our method is superior to the SR and other baseline methods. Conclusion: This paper proposed a new S-PET prediction method, which can significantly improve the PET image quality with low-dose injection. Significance: The proposed method is favorable in clinical application since it can decrease the potential radiation risk for patients.

Original languageEnglish
Article number7469380
Pages (from-to)569-579
Number of pages11
JournalIEEE Transactions on Biomedical Engineering
Volume64
Issue number3
DOIs
Publication statusPublished - 2017 Mar 1

Fingerprint

Positron emission tomography
Magnetic resonance
Glossaries
Positron-Emission Tomography
Magnetic Resonance Imaging
Learning
Imaging techniques
Image quality
Brain
Injections
Radiation

Keywords

  • Local coordinate coding (LCC)
  • positron emission tomography (PET)
  • semisupervised tripled dictionary learning (SSTDL)
  • sparse representation (SR)

ASJC Scopus subject areas

  • Biomedical Engineering

Cite this

Semisupervised Tripled Dictionary Learning for Standard-Dose PET Image Prediction Using Low-Dose PET and Multimodal MRI. / Wang, Yan; Ma, Guangkai; An, Le; Shi, Feng; Zhang, Pei; Lalush, David S.; Wu, Xi; Pu, Yifei; Zhou, Jiliu; Shen, Dinggang.

In: IEEE Transactions on Biomedical Engineering, Vol. 64, No. 3, 7469380, 01.03.2017, p. 569-579.

Research output: Contribution to journalArticle

Wang, Yan ; Ma, Guangkai ; An, Le ; Shi, Feng ; Zhang, Pei ; Lalush, David S. ; Wu, Xi ; Pu, Yifei ; Zhou, Jiliu ; Shen, Dinggang. / Semisupervised Tripled Dictionary Learning for Standard-Dose PET Image Prediction Using Low-Dose PET and Multimodal MRI. In: IEEE Transactions on Biomedical Engineering. 2017 ; Vol. 64, No. 3. pp. 569-579.
@article{3cd067f8cead44df861994cca3ab44ea,
title = "Semisupervised Tripled Dictionary Learning for Standard-Dose PET Image Prediction Using Low-Dose PET and Multimodal MRI",
abstract = "Objective: To obtain high-quality positron emission tomography (PET) image with low-dose tracer injection, this study attempts to predict the standard-dose PET (S-PET) image from both its low-dose PET (L-PET) counterpart and corresponding magnetic resonance imaging (MRI). Methods: It was achieved by patch-based sparse representation (SR), using the training samples with a complete set of MRI, L-PET and S-PET modalities for dictionary construction. However, the number of training samples with complete modalities is often limited. In practice, many samples generally have incomplete modalities (i.e., with one or two missing modalities) that thus cannot be used in the prediction process. In light of this, we develop a semisupervised tripled dictionary learning (SSTDL) method for S-PET image prediction, which can utilize not only the samples with complete modalities (called complete samples) but also the samples with incomplete modalities (called incomplete samples), to take advantage of the large number of available training samples and thus further improve the prediction performance. Results: Validation was done on a real human brain dataset consisting of 18 subjects, and the results show that our method is superior to the SR and other baseline methods. Conclusion: This paper proposed a new S-PET prediction method, which can significantly improve the PET image quality with low-dose injection. Significance: The proposed method is favorable in clinical application since it can decrease the potential radiation risk for patients.",
keywords = "Local coordinate coding (LCC), positron emission tomography (PET), semisupervised tripled dictionary learning (SSTDL), sparse representation (SR)",
author = "Yan Wang and Guangkai Ma and Le An and Feng Shi and Pei Zhang and Lalush, {David S.} and Xi Wu and Yifei Pu and Jiliu Zhou and Dinggang Shen",
year = "2017",
month = "3",
day = "1",
doi = "10.1109/TBME.2016.2564440",
language = "English",
volume = "64",
pages = "569--579",
journal = "IEEE Transactions on Biomedical Engineering",
issn = "0018-9294",
publisher = "IEEE Computer Society",
number = "3",

}

TY - JOUR

T1 - Semisupervised Tripled Dictionary Learning for Standard-Dose PET Image Prediction Using Low-Dose PET and Multimodal MRI

AU - Wang, Yan

AU - Ma, Guangkai

AU - An, Le

AU - Shi, Feng

AU - Zhang, Pei

AU - Lalush, David S.

AU - Wu, Xi

AU - Pu, Yifei

AU - Zhou, Jiliu

AU - Shen, Dinggang

PY - 2017/3/1

Y1 - 2017/3/1

N2 - Objective: To obtain high-quality positron emission tomography (PET) image with low-dose tracer injection, this study attempts to predict the standard-dose PET (S-PET) image from both its low-dose PET (L-PET) counterpart and corresponding magnetic resonance imaging (MRI). Methods: It was achieved by patch-based sparse representation (SR), using the training samples with a complete set of MRI, L-PET and S-PET modalities for dictionary construction. However, the number of training samples with complete modalities is often limited. In practice, many samples generally have incomplete modalities (i.e., with one or two missing modalities) that thus cannot be used in the prediction process. In light of this, we develop a semisupervised tripled dictionary learning (SSTDL) method for S-PET image prediction, which can utilize not only the samples with complete modalities (called complete samples) but also the samples with incomplete modalities (called incomplete samples), to take advantage of the large number of available training samples and thus further improve the prediction performance. Results: Validation was done on a real human brain dataset consisting of 18 subjects, and the results show that our method is superior to the SR and other baseline methods. Conclusion: This paper proposed a new S-PET prediction method, which can significantly improve the PET image quality with low-dose injection. Significance: The proposed method is favorable in clinical application since it can decrease the potential radiation risk for patients.

AB - Objective: To obtain high-quality positron emission tomography (PET) image with low-dose tracer injection, this study attempts to predict the standard-dose PET (S-PET) image from both its low-dose PET (L-PET) counterpart and corresponding magnetic resonance imaging (MRI). Methods: It was achieved by patch-based sparse representation (SR), using the training samples with a complete set of MRI, L-PET and S-PET modalities for dictionary construction. However, the number of training samples with complete modalities is often limited. In practice, many samples generally have incomplete modalities (i.e., with one or two missing modalities) that thus cannot be used in the prediction process. In light of this, we develop a semisupervised tripled dictionary learning (SSTDL) method for S-PET image prediction, which can utilize not only the samples with complete modalities (called complete samples) but also the samples with incomplete modalities (called incomplete samples), to take advantage of the large number of available training samples and thus further improve the prediction performance. Results: Validation was done on a real human brain dataset consisting of 18 subjects, and the results show that our method is superior to the SR and other baseline methods. Conclusion: This paper proposed a new S-PET prediction method, which can significantly improve the PET image quality with low-dose injection. Significance: The proposed method is favorable in clinical application since it can decrease the potential radiation risk for patients.

KW - Local coordinate coding (LCC)

KW - positron emission tomography (PET)

KW - semisupervised tripled dictionary learning (SSTDL)

KW - sparse representation (SR)

UR - http://www.scopus.com/inward/record.url?scp=85013420116&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85013420116&partnerID=8YFLogxK

U2 - 10.1109/TBME.2016.2564440

DO - 10.1109/TBME.2016.2564440

M3 - Article

C2 - 27187939

AN - SCOPUS:85013420116

VL - 64

SP - 569

EP - 579

JO - IEEE Transactions on Biomedical Engineering

JF - IEEE Transactions on Biomedical Engineering

SN - 0018-9294

IS - 3

M1 - 7469380

ER -