Validation of Deep-Learning Image Reconstruction for Low-Dose Chest Computed Tomography Scan: Emphasis on Image Quality and Noise

Objective: Iterative reconstruction degrades image quality. Thus, further advances in image reconstruction are necessary to overcome some limitations of this technique in low-dose computed tomography (LDCT) scan of the chest. Deep-learning image reconstruction (DLIR) is a new method used to reduce d...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Korean journal of radiology 2021, Vol.22 (1), p.131-138
Hauptverfasser: Joo Hee Kim, Hyun Jung Yoon, Eunju Lee, Injoong Kim, Yoon Ki Cha, So Hyeon Bak
Format: Artikel
Sprache:kor
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page 138
container_issue 1
container_start_page 131
container_title Korean journal of radiology
container_volume 22
creator Joo Hee Kim
Hyun Jung Yoon
Eunju Lee
Injoong Kim
Yoon Ki Cha
So Hyeon Bak
description Objective: Iterative reconstruction degrades image quality. Thus, further advances in image reconstruction are necessary to overcome some limitations of this technique in low-dose computed tomography (LDCT) scan of the chest. Deep-learning image reconstruction (DLIR) is a new method used to reduce dose while maintaining image quality. The purposes of this study was to evaluate image quality and noise of LDCT scan images reconstructed with DLIR and compare with those of images reconstructed with the adaptive statistical iterative reconstruction-Veo at a level of 30% (ASiR-V 30%). Materials and Methods: This retrospective study included 58 patients who underwent LDCT scan for lung cancer screening. Datasets were reconstructed with ASiR-V 30% and DLIR at medium and high levels (DLIR-M and DLIR-H, respectively). The objective image signal and noise, which represented mean attenuation value and standard deviation in Hounsfield units for the lungs, mediastinum, liver, and background air, and subjective image contrast, image noise, and conspicuity of structures were evaluated. The differences between CT scan images subjected to ASiR-V 30%, DLIR-M, and DLIR-H were evaluated. Results: Based on the objective analysis, the image signals did not significantly differ among ASiR-V 30%, DLIR-M, and DLIR-H (p = 0.949, 0.737, 0.366, and 0.358 in the lungs, mediastinum, liver, and background air, respectively). However, the noise was significantly lower in DLIR-M and DLIR-H than in ASiR-V 30% (all p < 0.001). DLIR had higher signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) than ASiR-V 30% (p = 0.027, < 0.001, and < 0.001 in the SNR of the lungs, mediastinum, and liver, respectively; all p < 0.001 in the CNR). According to the subjective analysis, DLIR had higher image contrast and lower image noise than ASiR-V 30% (all p < 0.001). DLIR was superior to ASiR-V 30% in identifying the pulmonary arteries and veins, trachea and bronchi, lymph nodes, and pleura and pericardium (all p < 0.001). Conclusion: DLIR significantly reduced the image noise in chest LDCT scan images compared with ASiR-V 30% while maintaining superior image quality.
format Article
fullrecord <record><control><sourceid>kisti</sourceid><recordid>TN_cdi_kisti_ndsl_JAKO202119077038176</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>JAKO202119077038176</sourcerecordid><originalsourceid>FETCH-kisti_ndsl_JAKO2021190770381763</originalsourceid><addsrcrecordid>eNqNjcFKw0AURQdRMFb_4W1cDkwm2mTcSdqibamoxW0ZkpdkbDIv5E2QLv1zS_UDXJ3NOfeeiUgrdS-zJFHnIoq1NnJqtLkUV8yfSmmjsrtIfH_Y1pU2OPJAFcwQe7lGO3jna3jubI3whgV5DsNYnKyKBljTl5wRI-QNcoCcun4MWMKWOqoH2zcHeC-sf4B51zeWHcMx_F17HY-H4QDWl7Ahx3gtLirbMt78cSJuF_Nt_iT3joPb-ZLb3fJx9aKVjmOj0lQlWZxOk_96Px21T5I</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Validation of Deep-Learning Image Reconstruction for Low-Dose Chest Computed Tomography Scan: Emphasis on Image Quality and Noise</title><source>KoreaMed Synapse</source><source>Elektronische Zeitschriftenbibliothek - Frei zugängliche E-Journals</source><source>PubMed Central Open Access</source><source>KoreaMed Open Access</source><source>PubMed Central</source><source>Alma/SFX Local Collection</source><creator>Joo Hee Kim ; Hyun Jung Yoon ; Eunju Lee ; Injoong Kim ; Yoon Ki Cha ; So Hyeon Bak</creator><creatorcontrib>Joo Hee Kim ; Hyun Jung Yoon ; Eunju Lee ; Injoong Kim ; Yoon Ki Cha ; So Hyeon Bak</creatorcontrib><description><![CDATA[Objective: Iterative reconstruction degrades image quality. Thus, further advances in image reconstruction are necessary to overcome some limitations of this technique in low-dose computed tomography (LDCT) scan of the chest. Deep-learning image reconstruction (DLIR) is a new method used to reduce dose while maintaining image quality. The purposes of this study was to evaluate image quality and noise of LDCT scan images reconstructed with DLIR and compare with those of images reconstructed with the adaptive statistical iterative reconstruction-Veo at a level of 30% (ASiR-V 30%). Materials and Methods: This retrospective study included 58 patients who underwent LDCT scan for lung cancer screening. Datasets were reconstructed with ASiR-V 30% and DLIR at medium and high levels (DLIR-M and DLIR-H, respectively). The objective image signal and noise, which represented mean attenuation value and standard deviation in Hounsfield units for the lungs, mediastinum, liver, and background air, and subjective image contrast, image noise, and conspicuity of structures were evaluated. The differences between CT scan images subjected to ASiR-V 30%, DLIR-M, and DLIR-H were evaluated. Results: Based on the objective analysis, the image signals did not significantly differ among ASiR-V 30%, DLIR-M, and DLIR-H (p = 0.949, 0.737, 0.366, and 0.358 in the lungs, mediastinum, liver, and background air, respectively). However, the noise was significantly lower in DLIR-M and DLIR-H than in ASiR-V 30% (all p < 0.001). DLIR had higher signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) than ASiR-V 30% (p = 0.027, < 0.001, and < 0.001 in the SNR of the lungs, mediastinum, and liver, respectively; all p < 0.001 in the CNR). According to the subjective analysis, DLIR had higher image contrast and lower image noise than ASiR-V 30% (all p < 0.001). DLIR was superior to ASiR-V 30% in identifying the pulmonary arteries and veins, trachea and bronchi, lymph nodes, and pleura and pericardium (all p < 0.001). Conclusion: DLIR significantly reduced the image noise in chest LDCT scan images compared with ASiR-V 30% while maintaining superior image quality.]]></description><identifier>ISSN: 1229-6929</identifier><identifier>EISSN: 2005-8330</identifier><language>kor</language><ispartof>Korean journal of radiology, 2021, Vol.22 (1), p.131-138</ispartof><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>230,314,780,784,885,4024</link.rule.ids></links><search><creatorcontrib>Joo Hee Kim</creatorcontrib><creatorcontrib>Hyun Jung Yoon</creatorcontrib><creatorcontrib>Eunju Lee</creatorcontrib><creatorcontrib>Injoong Kim</creatorcontrib><creatorcontrib>Yoon Ki Cha</creatorcontrib><creatorcontrib>So Hyeon Bak</creatorcontrib><title>Validation of Deep-Learning Image Reconstruction for Low-Dose Chest Computed Tomography Scan: Emphasis on Image Quality and Noise</title><title>Korean journal of radiology</title><addtitle>Korean journal of radiology : official journal of the Korean Radiological Society</addtitle><description><![CDATA[Objective: Iterative reconstruction degrades image quality. Thus, further advances in image reconstruction are necessary to overcome some limitations of this technique in low-dose computed tomography (LDCT) scan of the chest. Deep-learning image reconstruction (DLIR) is a new method used to reduce dose while maintaining image quality. The purposes of this study was to evaluate image quality and noise of LDCT scan images reconstructed with DLIR and compare with those of images reconstructed with the adaptive statistical iterative reconstruction-Veo at a level of 30% (ASiR-V 30%). Materials and Methods: This retrospective study included 58 patients who underwent LDCT scan for lung cancer screening. Datasets were reconstructed with ASiR-V 30% and DLIR at medium and high levels (DLIR-M and DLIR-H, respectively). The objective image signal and noise, which represented mean attenuation value and standard deviation in Hounsfield units for the lungs, mediastinum, liver, and background air, and subjective image contrast, image noise, and conspicuity of structures were evaluated. The differences between CT scan images subjected to ASiR-V 30%, DLIR-M, and DLIR-H were evaluated. Results: Based on the objective analysis, the image signals did not significantly differ among ASiR-V 30%, DLIR-M, and DLIR-H (p = 0.949, 0.737, 0.366, and 0.358 in the lungs, mediastinum, liver, and background air, respectively). However, the noise was significantly lower in DLIR-M and DLIR-H than in ASiR-V 30% (all p < 0.001). DLIR had higher signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) than ASiR-V 30% (p = 0.027, < 0.001, and < 0.001 in the SNR of the lungs, mediastinum, and liver, respectively; all p < 0.001 in the CNR). According to the subjective analysis, DLIR had higher image contrast and lower image noise than ASiR-V 30% (all p < 0.001). DLIR was superior to ASiR-V 30% in identifying the pulmonary arteries and veins, trachea and bronchi, lymph nodes, and pleura and pericardium (all p < 0.001). Conclusion: DLIR significantly reduced the image noise in chest LDCT scan images compared with ASiR-V 30% while maintaining superior image quality.]]></description><issn>1229-6929</issn><issn>2005-8330</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><sourceid>JDI</sourceid><recordid>eNqNjcFKw0AURQdRMFb_4W1cDkwm2mTcSdqibamoxW0ZkpdkbDIv5E2QLv1zS_UDXJ3NOfeeiUgrdS-zJFHnIoq1NnJqtLkUV8yfSmmjsrtIfH_Y1pU2OPJAFcwQe7lGO3jna3jubI3whgV5DsNYnKyKBljTl5wRI-QNcoCcun4MWMKWOqoH2zcHeC-sf4B51zeWHcMx_F17HY-H4QDWl7Ahx3gtLirbMt78cSJuF_Nt_iT3joPb-ZLb3fJx9aKVjmOj0lQlWZxOk_96Px21T5I</recordid><startdate>2021</startdate><enddate>2021</enddate><creator>Joo Hee Kim</creator><creator>Hyun Jung Yoon</creator><creator>Eunju Lee</creator><creator>Injoong Kim</creator><creator>Yoon Ki Cha</creator><creator>So Hyeon Bak</creator><scope>JDI</scope></search><sort><creationdate>2021</creationdate><title>Validation of Deep-Learning Image Reconstruction for Low-Dose Chest Computed Tomography Scan: Emphasis on Image Quality and Noise</title><author>Joo Hee Kim ; Hyun Jung Yoon ; Eunju Lee ; Injoong Kim ; Yoon Ki Cha ; So Hyeon Bak</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-kisti_ndsl_JAKO2021190770381763</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>kor</language><creationdate>2021</creationdate><toplevel>online_resources</toplevel><creatorcontrib>Joo Hee Kim</creatorcontrib><creatorcontrib>Hyun Jung Yoon</creatorcontrib><creatorcontrib>Eunju Lee</creatorcontrib><creatorcontrib>Injoong Kim</creatorcontrib><creatorcontrib>Yoon Ki Cha</creatorcontrib><creatorcontrib>So Hyeon Bak</creatorcontrib><collection>KoreaScience</collection><jtitle>Korean journal of radiology</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Joo Hee Kim</au><au>Hyun Jung Yoon</au><au>Eunju Lee</au><au>Injoong Kim</au><au>Yoon Ki Cha</au><au>So Hyeon Bak</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Validation of Deep-Learning Image Reconstruction for Low-Dose Chest Computed Tomography Scan: Emphasis on Image Quality and Noise</atitle><jtitle>Korean journal of radiology</jtitle><addtitle>Korean journal of radiology : official journal of the Korean Radiological Society</addtitle><date>2021</date><risdate>2021</risdate><volume>22</volume><issue>1</issue><spage>131</spage><epage>138</epage><pages>131-138</pages><issn>1229-6929</issn><eissn>2005-8330</eissn><abstract><![CDATA[Objective: Iterative reconstruction degrades image quality. Thus, further advances in image reconstruction are necessary to overcome some limitations of this technique in low-dose computed tomography (LDCT) scan of the chest. Deep-learning image reconstruction (DLIR) is a new method used to reduce dose while maintaining image quality. The purposes of this study was to evaluate image quality and noise of LDCT scan images reconstructed with DLIR and compare with those of images reconstructed with the adaptive statistical iterative reconstruction-Veo at a level of 30% (ASiR-V 30%). Materials and Methods: This retrospective study included 58 patients who underwent LDCT scan for lung cancer screening. Datasets were reconstructed with ASiR-V 30% and DLIR at medium and high levels (DLIR-M and DLIR-H, respectively). The objective image signal and noise, which represented mean attenuation value and standard deviation in Hounsfield units for the lungs, mediastinum, liver, and background air, and subjective image contrast, image noise, and conspicuity of structures were evaluated. The differences between CT scan images subjected to ASiR-V 30%, DLIR-M, and DLIR-H were evaluated. Results: Based on the objective analysis, the image signals did not significantly differ among ASiR-V 30%, DLIR-M, and DLIR-H (p = 0.949, 0.737, 0.366, and 0.358 in the lungs, mediastinum, liver, and background air, respectively). However, the noise was significantly lower in DLIR-M and DLIR-H than in ASiR-V 30% (all p < 0.001). DLIR had higher signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) than ASiR-V 30% (p = 0.027, < 0.001, and < 0.001 in the SNR of the lungs, mediastinum, and liver, respectively; all p < 0.001 in the CNR). According to the subjective analysis, DLIR had higher image contrast and lower image noise than ASiR-V 30% (all p < 0.001). DLIR was superior to ASiR-V 30% in identifying the pulmonary arteries and veins, trachea and bronchi, lymph nodes, and pleura and pericardium (all p < 0.001). Conclusion: DLIR significantly reduced the image noise in chest LDCT scan images compared with ASiR-V 30% while maintaining superior image quality.]]></abstract><oa>free_for_read</oa></addata></record>
fulltext fulltext
identifier ISSN: 1229-6929
ispartof Korean journal of radiology, 2021, Vol.22 (1), p.131-138
issn 1229-6929
2005-8330
language kor
recordid cdi_kisti_ndsl_JAKO202119077038176
source KoreaMed Synapse; Elektronische Zeitschriftenbibliothek - Frei zugängliche E-Journals; PubMed Central Open Access; KoreaMed Open Access; PubMed Central; Alma/SFX Local Collection
title Validation of Deep-Learning Image Reconstruction for Low-Dose Chest Computed Tomography Scan: Emphasis on Image Quality and Noise
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-05T09%3A55%3A38IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-kisti&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Validation%20of%20Deep-Learning%20Image%20Reconstruction%20for%20Low-Dose%20Chest%20Computed%20Tomography%20Scan:%20Emphasis%20on%20Image%20Quality%20and%20Noise&rft.jtitle=Korean%20journal%20of%20radiology&rft.au=Joo%20Hee%20Kim&rft.date=2021&rft.volume=22&rft.issue=1&rft.spage=131&rft.epage=138&rft.pages=131-138&rft.issn=1229-6929&rft.eissn=2005-8330&rft_id=info:doi/&rft_dat=%3Ckisti%3EJAKO202119077038176%3C/kisti%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true