An image-based deep learning framework for individualizing radiotherapy dose
Radiotherapy continues to be delivered uniformly without consideration of individual tumor characteristics. To advance toward more precise treatments in radiotherapy, we queried the lung computed tomography (CT)-derived feature space to identify radiation sensitivity parameters that can predict trea...
Gespeichert in:
| Veröffentlicht in: | The Lancet. Digital health 2019-07, Vol.1 (3), p.e136-e147 |
|---|---|
| Hauptverfasser: | , , , , , , , , |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | e147 |
|---|---|
| container_issue | 3 |
| container_start_page | e136 |
| container_title | The Lancet. Digital health |
| container_volume | 1 |
| creator | Lou, Bin Doken, Semihcan Zhuang, Tingliang Wingerter, Danielle Gidwani, Mishka Mistry, Nilesh Ladic, Lance Kamen, Ali Abazeed, Mohamed E |
| description | Radiotherapy continues to be delivered uniformly without consideration of individual tumor characteristics. To advance toward more precise treatments in radiotherapy, we queried the lung computed tomography (CT)-derived feature space to identify radiation sensitivity parameters that can predict treatment failure and hence guide the individualization of radiotherapy dose.
We used a cohort-based registry of 849 patients with cancer in the lung treated with high dose radiotherapy using stereotactic body radiotherapy. We input pre-therapy lung CT images into a multi-task deep neural network, Deep Profiler, to generate an image fingerprint that primarily predicts time to event treatment outcomes and secondarily approximates classical radiomic features. We validated our findings in an independent study population (
= 95). Deep Profiler was combined with clinical variables to derive
Gray, an individualized dose that estimates treatment failure probability to be |
| doi_str_mv | 10.1016/S2589-7500(19)30058-5 |
| format | Article |
| fullrecord | <record><control><sourceid>pubmed</sourceid><recordid>TN_cdi_pubmedcentral_primary_oai_pubmedcentral_nih_gov_6708276</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>31448366</sourcerecordid><originalsourceid>FETCH-LOGICAL-p1116-3bde2c605b9d597ba7736d0f786fdefa14c3e4bdb5a1c6ab3dec66afa69477163</originalsourceid><addsrcrecordid>eNpVkEFLAzEUhIMgttT-BCVHPawmm83L7kUoxapQ8KCel5fNSxvd7i7ZtlJ_vRa16GlghvlghrEzKa6kkHD9lOq8SIwW4kIWl0oInSf6iA0P9oCN-_5VCJGmUhljTthAySzLFcCQzScNDytcUGKxJ8cdUcdrwtiEZsF9xBW9t_GN-zby0LiwDW6DdfjYpxFdaNdLitjtuGt7OmXHHuuexj86Yi-z2-fpfTJ_vHuYTuZJJ6WERFlHaQVC28Lpwlg0RoET3uTgHXmUWaUos85qlBWgVY4qAPQIRWaMBDViN9_cbmNX5Cpq1hHrsotfS-KubDGU_5MmLMtFuy3BiDw1e8D5X8Ch-fuL-gSFOmfS</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>An image-based deep learning framework for individualizing radiotherapy dose</title><source>MEDLINE</source><source>Directory of Open Access Journals</source><source>Alma/SFX Local Collection</source><source>EZB Electronic Journals Library</source><creator>Lou, Bin ; Doken, Semihcan ; Zhuang, Tingliang ; Wingerter, Danielle ; Gidwani, Mishka ; Mistry, Nilesh ; Ladic, Lance ; Kamen, Ali ; Abazeed, Mohamed E</creator><creatorcontrib>Lou, Bin ; Doken, Semihcan ; Zhuang, Tingliang ; Wingerter, Danielle ; Gidwani, Mishka ; Mistry, Nilesh ; Ladic, Lance ; Kamen, Ali ; Abazeed, Mohamed E</creatorcontrib><description>Radiotherapy continues to be delivered uniformly without consideration of individual tumor characteristics. To advance toward more precise treatments in radiotherapy, we queried the lung computed tomography (CT)-derived feature space to identify radiation sensitivity parameters that can predict treatment failure and hence guide the individualization of radiotherapy dose.
We used a cohort-based registry of 849 patients with cancer in the lung treated with high dose radiotherapy using stereotactic body radiotherapy. We input pre-therapy lung CT images into a multi-task deep neural network, Deep Profiler, to generate an image fingerprint that primarily predicts time to event treatment outcomes and secondarily approximates classical radiomic features. We validated our findings in an independent study population (
= 95). Deep Profiler was combined with clinical variables to derive
Gray, an individualized dose that estimates treatment failure probability to be <5%.
Radiation treatments in patients with high Deep Profiler scores fail at a significantly higher rate than in those with low scores. The 3-year cumulative incidences of local failure were 20.3% (95% CI: 16.0-24.9) and 5.7% (95% CI: 3.5-8.8), respectively. Deep Profiler independently predicted local failure (hazard ratio 1.65, 95% 1.02-2.66,
= 0.04). Models that included Deep Profiler and clinical variables predicted treatment failures with a concordance index of 0.72 (95% CI: 0.67-0.77), a significant improvement compared to classical radiomics or clinical variables alone (
= <0.001 and <0.001, respectively). Deep Profiler performed well in an external study population (
= 95), accurately predicting treatment failures across diverse clinical settings and CT scanner types (concordance index = 0.77 [95% CI: 0.69-0.92]).
Gray had a wide dose range (21.1-277 Gy, BED), suggested dose reductions in 23.3% of patients and can be safely delivered in the majority of cases.
Our results indicate that there are image-distinct subpopulations that have differential sensitivity to radiotherapy. The image-based deep learning framework proposed herein is the first opportunity to use medical images to individualize radiotherapy dose.</description><identifier>EISSN: 2589-7500</identifier><identifier>DOI: 10.1016/S2589-7500(19)30058-5</identifier><identifier>PMID: 31448366</identifier><language>eng</language><publisher>England</publisher><subject>Aged ; Aged, 80 and over ; Carcinoma, Non-Small-Cell Lung - diagnostic imaging ; Carcinoma, Non-Small-Cell Lung - radiotherapy ; Deep Learning ; Female ; Humans ; Male ; Radiation Dosage ; Radiosurgery</subject><ispartof>The Lancet. Digital health, 2019-07, Vol.1 (3), p.e136-e147</ispartof><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>230,314,780,784,864,885,27924,27925</link.rule.ids><backlink>$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/31448366$$D View this record in MEDLINE/PubMed$$Hfree_for_read</backlink></links><search><creatorcontrib>Lou, Bin</creatorcontrib><creatorcontrib>Doken, Semihcan</creatorcontrib><creatorcontrib>Zhuang, Tingliang</creatorcontrib><creatorcontrib>Wingerter, Danielle</creatorcontrib><creatorcontrib>Gidwani, Mishka</creatorcontrib><creatorcontrib>Mistry, Nilesh</creatorcontrib><creatorcontrib>Ladic, Lance</creatorcontrib><creatorcontrib>Kamen, Ali</creatorcontrib><creatorcontrib>Abazeed, Mohamed E</creatorcontrib><title>An image-based deep learning framework for individualizing radiotherapy dose</title><title>The Lancet. Digital health</title><addtitle>Lancet Digit Health</addtitle><description>Radiotherapy continues to be delivered uniformly without consideration of individual tumor characteristics. To advance toward more precise treatments in radiotherapy, we queried the lung computed tomography (CT)-derived feature space to identify radiation sensitivity parameters that can predict treatment failure and hence guide the individualization of radiotherapy dose.
We used a cohort-based registry of 849 patients with cancer in the lung treated with high dose radiotherapy using stereotactic body radiotherapy. We input pre-therapy lung CT images into a multi-task deep neural network, Deep Profiler, to generate an image fingerprint that primarily predicts time to event treatment outcomes and secondarily approximates classical radiomic features. We validated our findings in an independent study population (
= 95). Deep Profiler was combined with clinical variables to derive
Gray, an individualized dose that estimates treatment failure probability to be <5%.
Radiation treatments in patients with high Deep Profiler scores fail at a significantly higher rate than in those with low scores. The 3-year cumulative incidences of local failure were 20.3% (95% CI: 16.0-24.9) and 5.7% (95% CI: 3.5-8.8), respectively. Deep Profiler independently predicted local failure (hazard ratio 1.65, 95% 1.02-2.66,
= 0.04). Models that included Deep Profiler and clinical variables predicted treatment failures with a concordance index of 0.72 (95% CI: 0.67-0.77), a significant improvement compared to classical radiomics or clinical variables alone (
= <0.001 and <0.001, respectively). Deep Profiler performed well in an external study population (
= 95), accurately predicting treatment failures across diverse clinical settings and CT scanner types (concordance index = 0.77 [95% CI: 0.69-0.92]).
Gray had a wide dose range (21.1-277 Gy, BED), suggested dose reductions in 23.3% of patients and can be safely delivered in the majority of cases.
Our results indicate that there are image-distinct subpopulations that have differential sensitivity to radiotherapy. The image-based deep learning framework proposed herein is the first opportunity to use medical images to individualize radiotherapy dose.</description><subject>Aged</subject><subject>Aged, 80 and over</subject><subject>Carcinoma, Non-Small-Cell Lung - diagnostic imaging</subject><subject>Carcinoma, Non-Small-Cell Lung - radiotherapy</subject><subject>Deep Learning</subject><subject>Female</subject><subject>Humans</subject><subject>Male</subject><subject>Radiation Dosage</subject><subject>Radiosurgery</subject><issn>2589-7500</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2019</creationdate><recordtype>article</recordtype><sourceid>EIF</sourceid><recordid>eNpVkEFLAzEUhIMgttT-BCVHPawmm83L7kUoxapQ8KCel5fNSxvd7i7ZtlJ_vRa16GlghvlghrEzKa6kkHD9lOq8SIwW4kIWl0oInSf6iA0P9oCN-_5VCJGmUhljTthAySzLFcCQzScNDytcUGKxJ8cdUcdrwtiEZsF9xBW9t_GN-zby0LiwDW6DdfjYpxFdaNdLitjtuGt7OmXHHuuexj86Yi-z2-fpfTJ_vHuYTuZJJ6WERFlHaQVC28Lpwlg0RoET3uTgHXmUWaUos85qlBWgVY4qAPQIRWaMBDViN9_cbmNX5Cpq1hHrsotfS-KubDGU_5MmLMtFuy3BiDw1e8D5X8Ch-fuL-gSFOmfS</recordid><startdate>201907</startdate><enddate>201907</enddate><creator>Lou, Bin</creator><creator>Doken, Semihcan</creator><creator>Zhuang, Tingliang</creator><creator>Wingerter, Danielle</creator><creator>Gidwani, Mishka</creator><creator>Mistry, Nilesh</creator><creator>Ladic, Lance</creator><creator>Kamen, Ali</creator><creator>Abazeed, Mohamed E</creator><scope>CGR</scope><scope>CUY</scope><scope>CVF</scope><scope>ECM</scope><scope>EIF</scope><scope>NPM</scope><scope>5PM</scope></search><sort><creationdate>201907</creationdate><title>An image-based deep learning framework for individualizing radiotherapy dose</title><author>Lou, Bin ; Doken, Semihcan ; Zhuang, Tingliang ; Wingerter, Danielle ; Gidwani, Mishka ; Mistry, Nilesh ; Ladic, Lance ; Kamen, Ali ; Abazeed, Mohamed E</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-p1116-3bde2c605b9d597ba7736d0f786fdefa14c3e4bdb5a1c6ab3dec66afa69477163</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2019</creationdate><topic>Aged</topic><topic>Aged, 80 and over</topic><topic>Carcinoma, Non-Small-Cell Lung - diagnostic imaging</topic><topic>Carcinoma, Non-Small-Cell Lung - radiotherapy</topic><topic>Deep Learning</topic><topic>Female</topic><topic>Humans</topic><topic>Male</topic><topic>Radiation Dosage</topic><topic>Radiosurgery</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Lou, Bin</creatorcontrib><creatorcontrib>Doken, Semihcan</creatorcontrib><creatorcontrib>Zhuang, Tingliang</creatorcontrib><creatorcontrib>Wingerter, Danielle</creatorcontrib><creatorcontrib>Gidwani, Mishka</creatorcontrib><creatorcontrib>Mistry, Nilesh</creatorcontrib><creatorcontrib>Ladic, Lance</creatorcontrib><creatorcontrib>Kamen, Ali</creatorcontrib><creatorcontrib>Abazeed, Mohamed E</creatorcontrib><collection>Medline</collection><collection>MEDLINE</collection><collection>MEDLINE (Ovid)</collection><collection>MEDLINE</collection><collection>MEDLINE</collection><collection>PubMed</collection><collection>PubMed Central (Full Participant titles)</collection><jtitle>The Lancet. Digital health</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Lou, Bin</au><au>Doken, Semihcan</au><au>Zhuang, Tingliang</au><au>Wingerter, Danielle</au><au>Gidwani, Mishka</au><au>Mistry, Nilesh</au><au>Ladic, Lance</au><au>Kamen, Ali</au><au>Abazeed, Mohamed E</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>An image-based deep learning framework for individualizing radiotherapy dose</atitle><jtitle>The Lancet. Digital health</jtitle><addtitle>Lancet Digit Health</addtitle><date>2019-07</date><risdate>2019</risdate><volume>1</volume><issue>3</issue><spage>e136</spage><epage>e147</epage><pages>e136-e147</pages><eissn>2589-7500</eissn><abstract>Radiotherapy continues to be delivered uniformly without consideration of individual tumor characteristics. To advance toward more precise treatments in radiotherapy, we queried the lung computed tomography (CT)-derived feature space to identify radiation sensitivity parameters that can predict treatment failure and hence guide the individualization of radiotherapy dose.
We used a cohort-based registry of 849 patients with cancer in the lung treated with high dose radiotherapy using stereotactic body radiotherapy. We input pre-therapy lung CT images into a multi-task deep neural network, Deep Profiler, to generate an image fingerprint that primarily predicts time to event treatment outcomes and secondarily approximates classical radiomic features. We validated our findings in an independent study population (
= 95). Deep Profiler was combined with clinical variables to derive
Gray, an individualized dose that estimates treatment failure probability to be <5%.
Radiation treatments in patients with high Deep Profiler scores fail at a significantly higher rate than in those with low scores. The 3-year cumulative incidences of local failure were 20.3% (95% CI: 16.0-24.9) and 5.7% (95% CI: 3.5-8.8), respectively. Deep Profiler independently predicted local failure (hazard ratio 1.65, 95% 1.02-2.66,
= 0.04). Models that included Deep Profiler and clinical variables predicted treatment failures with a concordance index of 0.72 (95% CI: 0.67-0.77), a significant improvement compared to classical radiomics or clinical variables alone (
= <0.001 and <0.001, respectively). Deep Profiler performed well in an external study population (
= 95), accurately predicting treatment failures across diverse clinical settings and CT scanner types (concordance index = 0.77 [95% CI: 0.69-0.92]).
Gray had a wide dose range (21.1-277 Gy, BED), suggested dose reductions in 23.3% of patients and can be safely delivered in the majority of cases.
Our results indicate that there are image-distinct subpopulations that have differential sensitivity to radiotherapy. The image-based deep learning framework proposed herein is the first opportunity to use medical images to individualize radiotherapy dose.</abstract><cop>England</cop><pmid>31448366</pmid><doi>10.1016/S2589-7500(19)30058-5</doi></addata></record> |
| fulltext | fulltext |
| identifier | EISSN: 2589-7500 |
| ispartof | The Lancet. Digital health, 2019-07, Vol.1 (3), p.e136-e147 |
| issn | 2589-7500 |
| language | eng |
| recordid | cdi_pubmedcentral_primary_oai_pubmedcentral_nih_gov_6708276 |
| source | MEDLINE; Directory of Open Access Journals; Alma/SFX Local Collection; EZB Electronic Journals Library |
| subjects | Aged Aged, 80 and over Carcinoma, Non-Small-Cell Lung - diagnostic imaging Carcinoma, Non-Small-Cell Lung - radiotherapy Deep Learning Female Humans Male Radiation Dosage Radiosurgery |
| title | An image-based deep learning framework for individualizing radiotherapy dose |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-20T17%3A15%3A15IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-pubmed&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=An%20image-based%20deep%20learning%20framework%20for%20individualizing%20radiotherapy%20dose&rft.jtitle=The%20Lancet.%20Digital%20health&rft.au=Lou,%20Bin&rft.date=2019-07&rft.volume=1&rft.issue=3&rft.spage=e136&rft.epage=e147&rft.pages=e136-e147&rft.eissn=2589-7500&rft_id=info:doi/10.1016/S2589-7500(19)30058-5&rft_dat=%3Cpubmed%3E31448366%3C/pubmed%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/31448366&rfr_iscdi=true |