Using radiomics based on multicenter magnetic resonance images to predict isocitrate dehydrogenase mutation status of gliomas
Isocitrate dehydrogenase (IDH) mutation status is an important biomarker for the treatment strategy selection and prognosis evaluation of glioma. The purpose of this study is to predict the IDH mutation status of gliomas based on multicenter magnetic resonance (MR) images using radiomic models, whic...
Gespeichert in:
Veröffentlicht in: | Quantitative imaging in medicine and surgery 2023-04, Vol.13 (4), p.2143-2155 |
---|---|
Hauptverfasser: | , , , , , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | 2155 |
---|---|
container_issue | 4 |
container_start_page | 2143 |
container_title | Quantitative imaging in medicine and surgery |
container_volume | 13 |
creator | Liu, Yan Zheng, Zhiming Wang, Zhiyuan Qian, Xusheng Yao, Zhigang Cheng, Chenchen Zhou, Zhiyong Gao, Fei Dai, Yakang |
description | Isocitrate dehydrogenase (IDH) mutation status is an important biomarker for the treatment strategy selection and prognosis evaluation of glioma. The purpose of this study is to predict the IDH mutation status of gliomas based on multicenter magnetic resonance (MR) images using radiomic models, which were composed from the selected radiomics features and logistic regression (LR), support vector machine (SVM), and LR least absolute shrinkage and selection operator (LASSO) classifiers.
We retrospectively reviewed the medical records of 205 patients with gliomas. We enrolled 78 patients from Shandong Provincial Hospital from January 2018 to December 2019 as testing sets and 127 patients from The Cancer Genome Atlas (TCGA) as training sets. Preoperative MR images were stratified according to their IDH status, and the participants formed a consecutive and random series. Four MR modalities, including T1C, T2, T1 fluid-attenuated inversion recovery (FLAIR), and T2 FLAIR, were used for analysis. Five-fold cross-validation was adopted to train the models, and the models' performances were verified through the testing set. Tumor volumes of interest (VOI) were delineated on the 4 MR modalities. A total of 428 radiomics features were extracted. Two feature selection algorithms, Pearson correlation coefficient (PCC) and recursive feature elimination (RFE), were used to select radiomics features. These features were fed into 3 machine learning classifiers, which were LR, SVM, and LR LASSO, to construct prediction models. The accuracy (ACC), sensitivity (SEN), specificity (SPEC), and area under the curve (AUC) were applied to measure the predictive performance of the radiomics models.
The LR (SVM and LR LASSO) classifier predicted IDH mutation status with an average testing set ACC of 80.77% (80.64% and 80.41%), a SEN of 73.68% (84.21% and 89.47%), a SPEC of 87.50% (67.50% and 62.50%), and an AUC of 0.8572 (0.8217 and 0.8164).
The radiomics models based on MR modalities demonstrated the potential to be used as tools across different data sets for the noninvasive prediction of the IDH mutation status in glioma. |
doi_str_mv | 10.21037/qims-22-836 |
format | Article |
fullrecord | <record><control><sourceid>proquest_pubme</sourceid><recordid>TN_cdi_pubmedcentral_primary_oai_pubmedcentral_nih_gov_10102787</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2802428403</sourcerecordid><originalsourceid>FETCH-LOGICAL-c385t-a31437b8a36bbad2354edded7735d2125e30e5d391727c38c298590a3d2a756f3</originalsourceid><addsrcrecordid>eNpVkc1PHiEQxkljU4311rPh2INbYdhd9j2ZxtjWxMRLPRMW5l1pduGVYU08-L8X60csl4GZh9_M5GHsixTfQAqlT-_CQg1AM6j-AzsAANW0SvR7r3fYwD47Ivoj6tGD1FJ8YvtKi75Vuj9gjzcU4sSz9SEtwREfLaHnKfJlnUtwGAtmvtgpYn3xjJSijQ55qDkkXhLfZfTBFR4ouVCyLcg93j74nCaMlVZJxZZQkVTjSjxt-TTXdpY-s49bOxMevcRDdvPj4vf5r-bq-ufl-ferxqmhK41Vsk47Dlb142g9qK5F79FrrToPEjpUAjuvNlKDrl8cbIZuI6zyYHXXb9UhO3vm7tZxQf-0Vbaz2eW6RX4wyQbzfyWGWzOleyOFFKAHXQlfXwg53a1IxSyBHM6zjZhWMjAIaGFoharSk2epy4ko4_atjxTmn2vmyTUDYKprVX78frY38atH6i9FsZca</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2802428403</pqid></control><display><type>article</type><title>Using radiomics based on multicenter magnetic resonance images to predict isocitrate dehydrogenase mutation status of gliomas</title><source>EZB-FREE-00999 freely available EZB journals</source><source>PubMed Central</source><creator>Liu, Yan ; Zheng, Zhiming ; Wang, Zhiyuan ; Qian, Xusheng ; Yao, Zhigang ; Cheng, Chenchen ; Zhou, Zhiyong ; Gao, Fei ; Dai, Yakang</creator><creatorcontrib>Liu, Yan ; Zheng, Zhiming ; Wang, Zhiyuan ; Qian, Xusheng ; Yao, Zhigang ; Cheng, Chenchen ; Zhou, Zhiyong ; Gao, Fei ; Dai, Yakang</creatorcontrib><description>Isocitrate dehydrogenase (IDH) mutation status is an important biomarker for the treatment strategy selection and prognosis evaluation of glioma. The purpose of this study is to predict the IDH mutation status of gliomas based on multicenter magnetic resonance (MR) images using radiomic models, which were composed from the selected radiomics features and logistic regression (LR), support vector machine (SVM), and LR least absolute shrinkage and selection operator (LASSO) classifiers.
We retrospectively reviewed the medical records of 205 patients with gliomas. We enrolled 78 patients from Shandong Provincial Hospital from January 2018 to December 2019 as testing sets and 127 patients from The Cancer Genome Atlas (TCGA) as training sets. Preoperative MR images were stratified according to their IDH status, and the participants formed a consecutive and random series. Four MR modalities, including T1C, T2, T1 fluid-attenuated inversion recovery (FLAIR), and T2 FLAIR, were used for analysis. Five-fold cross-validation was adopted to train the models, and the models' performances were verified through the testing set. Tumor volumes of interest (VOI) were delineated on the 4 MR modalities. A total of 428 radiomics features were extracted. Two feature selection algorithms, Pearson correlation coefficient (PCC) and recursive feature elimination (RFE), were used to select radiomics features. These features were fed into 3 machine learning classifiers, which were LR, SVM, and LR LASSO, to construct prediction models. The accuracy (ACC), sensitivity (SEN), specificity (SPEC), and area under the curve (AUC) were applied to measure the predictive performance of the radiomics models.
The LR (SVM and LR LASSO) classifier predicted IDH mutation status with an average testing set ACC of 80.77% (80.64% and 80.41%), a SEN of 73.68% (84.21% and 89.47%), a SPEC of 87.50% (67.50% and 62.50%), and an AUC of 0.8572 (0.8217 and 0.8164).
The radiomics models based on MR modalities demonstrated the potential to be used as tools across different data sets for the noninvasive prediction of the IDH mutation status in glioma.</description><identifier>ISSN: 2223-4292</identifier><identifier>EISSN: 2223-4306</identifier><identifier>DOI: 10.21037/qims-22-836</identifier><identifier>PMID: 37064376</identifier><language>eng</language><publisher>China: AME Publishing Company</publisher><subject>Original</subject><ispartof>Quantitative imaging in medicine and surgery, 2023-04, Vol.13 (4), p.2143-2155</ispartof><rights>2023 Quantitative Imaging in Medicine and Surgery. All rights reserved.</rights><rights>2023 Quantitative Imaging in Medicine and Surgery. All rights reserved. 2023 Quantitative Imaging in Medicine and Surgery.</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c385t-a31437b8a36bbad2354edded7735d2125e30e5d391727c38c298590a3d2a756f3</citedby></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktopdf>$$Uhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC10102787/pdf/$$EPDF$$P50$$Gpubmedcentral$$Hfree_for_read</linktopdf><linktohtml>$$Uhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC10102787/$$EHTML$$P50$$Gpubmedcentral$$Hfree_for_read</linktohtml><link.rule.ids>230,314,727,780,784,885,27924,27925,53791,53793</link.rule.ids><backlink>$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/37064376$$D View this record in MEDLINE/PubMed$$Hfree_for_read</backlink></links><search><creatorcontrib>Liu, Yan</creatorcontrib><creatorcontrib>Zheng, Zhiming</creatorcontrib><creatorcontrib>Wang, Zhiyuan</creatorcontrib><creatorcontrib>Qian, Xusheng</creatorcontrib><creatorcontrib>Yao, Zhigang</creatorcontrib><creatorcontrib>Cheng, Chenchen</creatorcontrib><creatorcontrib>Zhou, Zhiyong</creatorcontrib><creatorcontrib>Gao, Fei</creatorcontrib><creatorcontrib>Dai, Yakang</creatorcontrib><title>Using radiomics based on multicenter magnetic resonance images to predict isocitrate dehydrogenase mutation status of gliomas</title><title>Quantitative imaging in medicine and surgery</title><addtitle>Quant Imaging Med Surg</addtitle><description>Isocitrate dehydrogenase (IDH) mutation status is an important biomarker for the treatment strategy selection and prognosis evaluation of glioma. The purpose of this study is to predict the IDH mutation status of gliomas based on multicenter magnetic resonance (MR) images using radiomic models, which were composed from the selected radiomics features and logistic regression (LR), support vector machine (SVM), and LR least absolute shrinkage and selection operator (LASSO) classifiers.
We retrospectively reviewed the medical records of 205 patients with gliomas. We enrolled 78 patients from Shandong Provincial Hospital from January 2018 to December 2019 as testing sets and 127 patients from The Cancer Genome Atlas (TCGA) as training sets. Preoperative MR images were stratified according to their IDH status, and the participants formed a consecutive and random series. Four MR modalities, including T1C, T2, T1 fluid-attenuated inversion recovery (FLAIR), and T2 FLAIR, were used for analysis. Five-fold cross-validation was adopted to train the models, and the models' performances were verified through the testing set. Tumor volumes of interest (VOI) were delineated on the 4 MR modalities. A total of 428 radiomics features were extracted. Two feature selection algorithms, Pearson correlation coefficient (PCC) and recursive feature elimination (RFE), were used to select radiomics features. These features were fed into 3 machine learning classifiers, which were LR, SVM, and LR LASSO, to construct prediction models. The accuracy (ACC), sensitivity (SEN), specificity (SPEC), and area under the curve (AUC) were applied to measure the predictive performance of the radiomics models.
The LR (SVM and LR LASSO) classifier predicted IDH mutation status with an average testing set ACC of 80.77% (80.64% and 80.41%), a SEN of 73.68% (84.21% and 89.47%), a SPEC of 87.50% (67.50% and 62.50%), and an AUC of 0.8572 (0.8217 and 0.8164).
The radiomics models based on MR modalities demonstrated the potential to be used as tools across different data sets for the noninvasive prediction of the IDH mutation status in glioma.</description><subject>Original</subject><issn>2223-4292</issn><issn>2223-4306</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><recordid>eNpVkc1PHiEQxkljU4311rPh2INbYdhd9j2ZxtjWxMRLPRMW5l1pduGVYU08-L8X60csl4GZh9_M5GHsixTfQAqlT-_CQg1AM6j-AzsAANW0SvR7r3fYwD47Ivoj6tGD1FJ8YvtKi75Vuj9gjzcU4sSz9SEtwREfLaHnKfJlnUtwGAtmvtgpYn3xjJSijQ55qDkkXhLfZfTBFR4ouVCyLcg93j74nCaMlVZJxZZQkVTjSjxt-TTXdpY-s49bOxMevcRDdvPj4vf5r-bq-ufl-ferxqmhK41Vsk47Dlb142g9qK5F79FrrToPEjpUAjuvNlKDrl8cbIZuI6zyYHXXb9UhO3vm7tZxQf-0Vbaz2eW6RX4wyQbzfyWGWzOleyOFFKAHXQlfXwg53a1IxSyBHM6zjZhWMjAIaGFoharSk2epy4ko4_atjxTmn2vmyTUDYKprVX78frY38atH6i9FsZca</recordid><startdate>20230401</startdate><enddate>20230401</enddate><creator>Liu, Yan</creator><creator>Zheng, Zhiming</creator><creator>Wang, Zhiyuan</creator><creator>Qian, Xusheng</creator><creator>Yao, Zhigang</creator><creator>Cheng, Chenchen</creator><creator>Zhou, Zhiyong</creator><creator>Gao, Fei</creator><creator>Dai, Yakang</creator><general>AME Publishing Company</general><scope>NPM</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7X8</scope><scope>5PM</scope></search><sort><creationdate>20230401</creationdate><title>Using radiomics based on multicenter magnetic resonance images to predict isocitrate dehydrogenase mutation status of gliomas</title><author>Liu, Yan ; Zheng, Zhiming ; Wang, Zhiyuan ; Qian, Xusheng ; Yao, Zhigang ; Cheng, Chenchen ; Zhou, Zhiyong ; Gao, Fei ; Dai, Yakang</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c385t-a31437b8a36bbad2354edded7735d2125e30e5d391727c38c298590a3d2a756f3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Original</topic><toplevel>online_resources</toplevel><creatorcontrib>Liu, Yan</creatorcontrib><creatorcontrib>Zheng, Zhiming</creatorcontrib><creatorcontrib>Wang, Zhiyuan</creatorcontrib><creatorcontrib>Qian, Xusheng</creatorcontrib><creatorcontrib>Yao, Zhigang</creatorcontrib><creatorcontrib>Cheng, Chenchen</creatorcontrib><creatorcontrib>Zhou, Zhiyong</creatorcontrib><creatorcontrib>Gao, Fei</creatorcontrib><creatorcontrib>Dai, Yakang</creatorcontrib><collection>PubMed</collection><collection>CrossRef</collection><collection>MEDLINE - Academic</collection><collection>PubMed Central (Full Participant titles)</collection><jtitle>Quantitative imaging in medicine and surgery</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Liu, Yan</au><au>Zheng, Zhiming</au><au>Wang, Zhiyuan</au><au>Qian, Xusheng</au><au>Yao, Zhigang</au><au>Cheng, Chenchen</au><au>Zhou, Zhiyong</au><au>Gao, Fei</au><au>Dai, Yakang</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Using radiomics based on multicenter magnetic resonance images to predict isocitrate dehydrogenase mutation status of gliomas</atitle><jtitle>Quantitative imaging in medicine and surgery</jtitle><addtitle>Quant Imaging Med Surg</addtitle><date>2023-04-01</date><risdate>2023</risdate><volume>13</volume><issue>4</issue><spage>2143</spage><epage>2155</epage><pages>2143-2155</pages><issn>2223-4292</issn><eissn>2223-4306</eissn><abstract>Isocitrate dehydrogenase (IDH) mutation status is an important biomarker for the treatment strategy selection and prognosis evaluation of glioma. The purpose of this study is to predict the IDH mutation status of gliomas based on multicenter magnetic resonance (MR) images using radiomic models, which were composed from the selected radiomics features and logistic regression (LR), support vector machine (SVM), and LR least absolute shrinkage and selection operator (LASSO) classifiers.
We retrospectively reviewed the medical records of 205 patients with gliomas. We enrolled 78 patients from Shandong Provincial Hospital from January 2018 to December 2019 as testing sets and 127 patients from The Cancer Genome Atlas (TCGA) as training sets. Preoperative MR images were stratified according to their IDH status, and the participants formed a consecutive and random series. Four MR modalities, including T1C, T2, T1 fluid-attenuated inversion recovery (FLAIR), and T2 FLAIR, were used for analysis. Five-fold cross-validation was adopted to train the models, and the models' performances were verified through the testing set. Tumor volumes of interest (VOI) were delineated on the 4 MR modalities. A total of 428 radiomics features were extracted. Two feature selection algorithms, Pearson correlation coefficient (PCC) and recursive feature elimination (RFE), were used to select radiomics features. These features were fed into 3 machine learning classifiers, which were LR, SVM, and LR LASSO, to construct prediction models. The accuracy (ACC), sensitivity (SEN), specificity (SPEC), and area under the curve (AUC) were applied to measure the predictive performance of the radiomics models.
The LR (SVM and LR LASSO) classifier predicted IDH mutation status with an average testing set ACC of 80.77% (80.64% and 80.41%), a SEN of 73.68% (84.21% and 89.47%), a SPEC of 87.50% (67.50% and 62.50%), and an AUC of 0.8572 (0.8217 and 0.8164).
The radiomics models based on MR modalities demonstrated the potential to be used as tools across different data sets for the noninvasive prediction of the IDH mutation status in glioma.</abstract><cop>China</cop><pub>AME Publishing Company</pub><pmid>37064376</pmid><doi>10.21037/qims-22-836</doi><tpages>13</tpages><oa>free_for_read</oa></addata></record> |
fulltext | fulltext |
identifier | ISSN: 2223-4292 |
ispartof | Quantitative imaging in medicine and surgery, 2023-04, Vol.13 (4), p.2143-2155 |
issn | 2223-4292 2223-4306 |
language | eng |
recordid | cdi_pubmedcentral_primary_oai_pubmedcentral_nih_gov_10102787 |
source | EZB-FREE-00999 freely available EZB journals; PubMed Central |
subjects | Original |
title | Using radiomics based on multicenter magnetic resonance images to predict isocitrate dehydrogenase mutation status of gliomas |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-25T17%3A46%3A21IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_pubme&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Using%20radiomics%20based%20on%20multicenter%20magnetic%20resonance%20images%20to%20predict%20isocitrate%20dehydrogenase%20mutation%20status%20of%20gliomas&rft.jtitle=Quantitative%20imaging%20in%20medicine%20and%20surgery&rft.au=Liu,%20Yan&rft.date=2023-04-01&rft.volume=13&rft.issue=4&rft.spage=2143&rft.epage=2155&rft.pages=2143-2155&rft.issn=2223-4292&rft.eissn=2223-4306&rft_id=info:doi/10.21037/qims-22-836&rft_dat=%3Cproquest_pubme%3E2802428403%3C/proquest_pubme%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2802428403&rft_id=info:pmid/37064376&rfr_iscdi=true |