Gradient-based Optimization Algorithm for Hybrid Loss Function in Low-dose CT Denoising

Deep learning techniques have emerged in de-noising low-dose computed tomography (CT) images to avoid the potential health risks of high ionizing radiation dose on patients. Although these post-processing methods display high quality denoised images, the denoising performance still has the potential...

Full description

Saved in:
Bibliographic Details
Published in:2022 44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC) Vol. 2022; pp. 3834 - 3838
Main Authors: Mazandarani, Farzan Niknejad, Marcos, Luella, Babyn, Paul, Alirezaie, Javad
Format: Conference Proceeding Journal Article
Language:English
Published: IEEE 01.01.2022
Subjects:
ISSN:2694-0604, 2694-0604
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract Deep learning techniques have emerged in de-noising low-dose computed tomography (CT) images to avoid the potential health risks of high ionizing radiation dose on patients. Although these post-processing methods display high quality denoised images, the denoising performance still has the potential to improve. The primary purpose of this work was to determine and analyze the most effective and efficient hybrid loss function in deep learning (DL)-based denoising network. Objective functions in deep learning algorithms are the main keys for optimizing the parameters of a network and can affect the quality of the denoised image significantly. Hence, this work examined the various combinations of the most common objective functions in CT denoising networks, namely L1 loss, per-pixel loss, perceptual loss, and structural dissimilarity loss. Further, a hyperparameter learning algorithm was also introduced to find the best scalable factors of the loss functions in each hybrid loss function combination. For simplic-ity, RED-CNN was used in this study to easily demonstrate the performance of the losses during the denoising process. Based on this experiment, the balance between these loss function via the gradient-based optimization algorithm could help in the generalizability prediction of designing future CT denoising networks.
AbstractList Deep learning techniques have emerged in de-noising low-dose computed tomography (CT) images to avoid the potential health risks of high ionizing radiation dose on patients. Although these post-processing methods display high quality denoised images, the denoising performance still has the potential to improve. The primary purpose of this work was to determine and analyze the most effective and efficient hybrid loss function in deep learning (DL)-based denoising network. Objective functions in deep learning algorithms are the main keys for optimizing the parameters of a network and can affect the quality of the denoised image significantly. Hence, this work examined the various combinations of the most common objective functions in CT denoising networks, namely L1 loss, per-pixel loss, perceptual loss, and structural dissimilarity loss. Further, a hyperparameter learning algorithm was also introduced to find the best scalable factors of the loss functions in each hybrid loss function combination. For simplic-ity, RED-CNN was used in this study to easily demonstrate the performance of the losses during the denoising process. Based on this experiment, the balance between these loss function via the gradient-based optimization algorithm could help in the generalizability prediction of designing future CT denoising networks.
Deep learning techniques have emerged in de-noising low-dose computed tomography (CT) images to avoid the potential health risks of high ionizing radiation dose on patients. Although these post-processing methods display high quality denoised images, the denoising performance still has the potential to improve. The primary purpose of this work was to determine and analyze the most effective and efficient hybrid loss function in deep learning (DL)-based denoising network. Objective functions in deep learning algorithms are the main keys for optimizing the parameters of a network and can affect the quality of the denoised image significantly. Hence, this work examined the various combinations of the most common objective functions in CT denoising networks, namely L1 loss, per-pixel loss, perceptual loss, and structural dissimilarity loss. Further, a hyperparameter learning algorithm was also introduced to find the best scalable factors of the loss functions in each hybrid loss function combination. For simplic-ity, RED-CNN was used in this study to easily demonstrate the performance of the losses during the denoising process. Based on this experiment, the balance between these loss function via the gradient-based optimization algorithm could help in the generalizability prediction of designing future CT denoising networks.Deep learning techniques have emerged in de-noising low-dose computed tomography (CT) images to avoid the potential health risks of high ionizing radiation dose on patients. Although these post-processing methods display high quality denoised images, the denoising performance still has the potential to improve. The primary purpose of this work was to determine and analyze the most effective and efficient hybrid loss function in deep learning (DL)-based denoising network. Objective functions in deep learning algorithms are the main keys for optimizing the parameters of a network and can affect the quality of the denoised image significantly. Hence, this work examined the various combinations of the most common objective functions in CT denoising networks, namely L1 loss, per-pixel loss, perceptual loss, and structural dissimilarity loss. Further, a hyperparameter learning algorithm was also introduced to find the best scalable factors of the loss functions in each hybrid loss function combination. For simplic-ity, RED-CNN was used in this study to easily demonstrate the performance of the losses during the denoising process. Based on this experiment, the balance between these loss function via the gradient-based optimization algorithm could help in the generalizability prediction of designing future CT denoising networks.
Author Babyn, Paul
Alirezaie, Javad
Mazandarani, Farzan Niknejad
Marcos, Luella
Author_xml – sequence: 1
  givenname: Farzan Niknejad
  surname: Mazandarani
  fullname: Mazandarani, Farzan Niknejad
  email: fniknejad@ryerson.ca
  organization: Ryerson University,Department of Electrical, Computer and Biomedical Engineering,Toronto,ON,Canada,M5B2K3
– sequence: 2
  givenname: Luella
  surname: Marcos
  fullname: Marcos, Luella
  email: lgmarcos@ryerson.ca
  organization: Ryerson University,Department of Electrical, Computer and Biomedical Engineering,Toronto,ON,Canada,M5B2K3
– sequence: 3
  givenname: Paul
  surname: Babyn
  fullname: Babyn, Paul
  email: paul.babyn@saskatoonhealthregion.ca
  organization: University of Saskatoon Health Region, Royal University Hospital,Department of Medical Imaging,Saskatoon,SK,Canada,S7NOW8
– sequence: 4
  givenname: Javad
  surname: Alirezaie
  fullname: Alirezaie, Javad
  email: javad@ryerson.ca
  organization: Ryerson University,Department of Electrical, Computer and Biomedical Engineering,Toronto,ON,Canada,M5B2K3
BookMark eNpNkMFOwzAQRA0qEm3pFyAhH7mk2Bs7cY4ltBSpqJcijpGdbIqlxi5xKlS-nkB74LS7o6ed0YzIwHmHhNxxNuWcZQ_z18dcKIBsCgxgmqmUx4pdkBFPQXFIFcSXZAhJJiKWMDH4t1-TSQjWMBlLITOIh-T9udWVRddFRges6Hrf2cZ-6856R2e7rW9t99HQ2rd0eTStrejKh0AXB1f-Idb1wldU-YA039AndN4G67Y35KrWu4CT8xyTt8V8ky-j1fr5JZ-tIgtx0psywfrwRmLJ6lTISiQ1R50pA7IHElkyBbXRsUCQxiQlZgbK_qgMclRJPCb3p7_71n8eMHRFY0OJu5126A-hgJSD-m1L9ujtCbWIWOxb2-j2WJzri38AVMxkSQ
ContentType Conference Proceeding
Journal Article
DBID 6IE
6IH
CBEJK
RIE
RIO
7X8
DOI 10.1109/EMBC48229.2022.9871380
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Proceedings Order Plan (POP) 1998-present by volume
IEEE Xplore All Conference Proceedings
IEEE/IET Electronic Library (IEL)
IEEE Proceedings Order Plans (POP) 1998-present
MEDLINE - Academic
DatabaseTitle MEDLINE - Academic
DatabaseTitleList
MEDLINE - Academic
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE/IET Electronic Library (IEL)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
– sequence: 2
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod fulltext_linktorsrc
EISBN 1728127823
9781728127828
EISSN 2694-0604
EndPage 3838
ExternalDocumentID 9871380
Genre orig-research
GroupedDBID 6IE
6IH
CBEJK
RIE
RIO
7X8
ID FETCH-LOGICAL-i236t-b040380b5ec0f745d46f1ea98b2523665c082fba34e25bb6ce9b2c4e2dbe1e863
IEDL.DBID RIE
ISICitedReferencesCount 8
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001251147203147&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 2694-0604
IngestDate Wed Oct 01 13:48:43 EDT 2025
Wed Aug 27 02:19:01 EDT 2025
IsPeerReviewed true
IsScholarly false
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-i236t-b040380b5ec0f745d46f1ea98b2523665c082fba34e25bb6ce9b2c4e2dbe1e863
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
PQID 2712848225
PQPubID 23479
PageCount 5
ParticipantIDs proquest_miscellaneous_2712848225
ieee_primary_9871380
PublicationCentury 2000
PublicationDate 2022-01-01
PublicationDateYYYYMMDD 2022-01-01
PublicationDate_xml – month: 01
  year: 2022
  text: 2022-01-01
  day: 01
PublicationDecade 2020
PublicationTitle 2022 44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC)
PublicationTitleAbbrev EMBC
PublicationYear 2022
Publisher IEEE
Publisher_xml – name: IEEE
SSID ssib053545923
ssib042469959
ssib058575407
ssib061542107
Score 1.8361275
Snippet Deep learning techniques have emerged in de-noising low-dose computed tomography (CT) images to avoid the potential health risks of high ionizing radiation...
SourceID proquest
ieee
SourceType Aggregation Database
Publisher
StartPage 3834
SubjectTerms Biological system modeling
Computed tomography
CT image denoising
Deep learning
Ionizing radiation
Loss functions
Medical imaging
Noise reduction
opti-mization algorithm
Prediction algorithms
RED-CNN
Signal Processing
Three-dimensional displays
Title Gradient-based Optimization Algorithm for Hybrid Loss Function in Low-dose CT Denoising
URI https://ieeexplore.ieee.org/document/9871380
https://www.proquest.com/docview/2712848225
Volume 2022
WOSCitedRecordID wos001251147203147&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LT8JAEN4g8eBJDRjxQdbEo4V2X22PiiAHRA4YuZHddqpNpDU8NP57Z0tREr14a5ttu5md3W--2Z0ZQi4l8z3txdpBcNOO4MJ1tOauw4zyIAkh5LrIrj_wh8NgMglHFXL1HQsDAMXhM2jZy2IvP86jlXWVtZEfezxAgr7j-2odq7XRHcGQ523lSZEcTYMtaiBtJUokL2WQsOeG7e79TUfYfOfIEhlrlR8vq6z8WpoLvOnt_6-nB6T-E7hHR9-QdEgqkNXI0928ONe1dCxkxfQBV4lZGX5Jr1-f83m6fJlRtF5p_9MGcNEBIiftIeIVTdIMH3w4cb4A2hnTW8jy1HoY6uSx1x13-k5ZT8FJGVf4E5yw2CkjIXITX8hYqMQDHQaGIR1VSkZoDyRGcwFMGqMiCA2L8CY24EGg-BGpZnkGx4QKJcB3LTvkRkRcaU8mJpAaZ3PCkyBskJoVyvRtnTJjWsqjQS42Up2iGtu9CZ1BvlpMUWUQKHE05Mnfr56SPTtMa-_HGaku5ys4J7vR-zJdzJuoEZOgWWjEF4XgsnY
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1LT8MwDI7QQIIToA3xJkgcKbR5tT3CYAxRxg5D7FYlrQuVWIv2APHvcUoZSHDh1lZpGzlOPn9ObBNyJJnvaS_VDoKbdgQXrqM1dx1mlAdZCCHXVXb9yO_1guEw7C-Q43ksDABUh8_gxF5We_lpmcysq-wU-bHHAyToi7ZyVh2t9aU9giHT-5EpRXI0Dn6QA2lrUSJ9qcOEPTc8vbw9bwub8Rx5ImMn9efrOiu_FucKcTqr_-vrGml9h-7R_hyU1skCFE3ycDWuTnZNHQtaKb3DdWJUB2DSs-fHcpxPn0YU7VfafbchXDRC7KQdxLyqSV7ggzcnLSdA2wN6AUWZWx9Di9x3LgftrlNXVHByxhX-BKcsdspISNzMFzIVKvNAh4FhSEiVkglaBJnRXACTxqgEQsMSvEkNeBAovkEaRVnAJqFCCfBdyw-5EQlX2pOZCaTG-ZzxLAi3SNMKJX75TJoR1_LYIodfUo1Rke3uhC6gnE1iVBqEShwNuf33qwdkuTu4jeLounezQ1bskH36QnZJYzqewR5ZSl6n-WS8X-nFB8pntNc
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=proceeding&rft.title=2022+44th+Annual+International+Conference+of+the+IEEE+Engineering+in+Medicine+%26+Biology+Society+%28EMBC%29&rft.atitle=Gradient-based+Optimization+Algorithm+for+Hybrid+Loss+Function+in+Low-dose+CT+Denoising&rft.au=Mazandarani%2C+Farzan+Niknejad&rft.au=Marcos%2C+Luella&rft.au=Babyn%2C+Paul&rft.au=Alirezaie%2C+Javad&rft.date=2022-01-01&rft.pub=IEEE&rft.eissn=2694-0604&rft.spage=3834&rft.epage=3838&rft_id=info:doi/10.1109%2FEMBC48229.2022.9871380&rft.externalDocID=9871380
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2694-0604&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2694-0604&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2694-0604&client=summon