Thyroid Ultrasound Reports: Will the Thyroid Imaging, Reporting, and Data System Improve Natural Language Processing Capture of Critical Thyroid Nodule Features?
Critical thyroid nodule features are contained in unstructured ultrasound (US) reports. The Thyroid Imaging, Reporting, and Data System (TI-RADS) uses five key features to risk stratify nodules and recommend appropriate intervention. This study aims to analyze the quality of US reporting and the pot...
Gespeichert in:
| Veröffentlicht in: | The Journal of surgical research 2020-12, Vol.256, p.557-563 |
|---|---|
| Hauptverfasser: | , , , |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 563 |
|---|---|
| container_issue | |
| container_start_page | 557 |
| container_title | The Journal of surgical research |
| container_volume | 256 |
| creator | Chen, Kallie J. Dedhia, Priya H. Imbus, Joseph R. Schneider, David F. |
| description | Critical thyroid nodule features are contained in unstructured ultrasound (US) reports. The Thyroid Imaging, Reporting, and Data System (TI-RADS) uses five key features to risk stratify nodules and recommend appropriate intervention. This study aims to analyze the quality of US reporting and the potential benefit of Natural Language Processing (NLP) systems in efficiently capturing TI-RADS features from text reports.
This retrospective study used free-text thyroid US reports from an academic center (A) and community hospital (B). Physicians created “gold standard” annotations by manually extracting TI-RADS features and clinical recommendations from reports to determine how often they were included. Similar annotations were created using an automated NLP system and compared with the gold standard.
Two hundred eighty-two reports contained 409 nodules at least 1-cm in maximum diameter. The gold standard identified three nodules (0.7%) which contained enough information to calculate a complete TI-RADS score. Shape was described most often (92.7% of nodules), whereas margins were described least often (11%). A median number of two TI-RADS features are reported per nodule. The NLP system was significantly less accurate than the gold standard in capturing echogenicity (27.5%) and margins (58.9%). One hundred eight nodule reports (26.4%) included clinical management recommendations, which were included more often at site A than B (33.9 versus 17%, P |
| doi_str_mv | 10.1016/j.jss.2020.07.015 |
| format | Article |
| fullrecord | <record><control><sourceid>proquest_cross</sourceid><recordid>TN_cdi_proquest_miscellaneous_2434755008</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><els_id>S0022480420304728</els_id><sourcerecordid>2434755008</sourcerecordid><originalsourceid>FETCH-LOGICAL-c326t-83b805a93ffe1a7b9d9fb5ad76bcdb9b7845a31bb0c51edcfe2ae9b024803beb3</originalsourceid><addsrcrecordid>eNp9kd1u0zAYhi0EYt3gAjhBPuSAZJ_jpEngYEJlY5OqMY1NHFr--dK5SuJiO5N6OdwpLm055Mi2_Lyvfx5C3jHIGbD5-Tpfh5AXUEAOdQ6sekFmDNoqa-Y1f0lmAEWRlQ2UJ-Q0hDWkdVvz1-SEF3XbAlQz8vvhaeudNfSxj14GN42G3uPG-Rg-0Z-272l8QnqEbga5suPq4wH5O5Up8VVGSX9sQ8QhMRvvnpHeyjh52dOlHFeTXCG9805jCClEF3KTNpG6ji68jVYn7njGrTNTj_QKd3kMF2_Iq072Ad8exjPyeHX5sLjOlt-_3Sy-LDPNi3nMGq4aqGTLuw6ZrFVr2k5V0tRzpY1qVd2UleRMKdAVQ6M7LCS2Cor0P1yh4mfkw743Xf_XhCGKwQaNfS9HdFMQRcnLuqoAmoSyPaq9C8FjJzbeDtJvBQOxMyPWIpkROzMCapHMpMz7Q_2kBjT_EkcVCfi8BzA98tmiF0FbHDUa61FHYZz9T_0fiSyibQ</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2434755008</pqid></control><display><type>article</type><title>Thyroid Ultrasound Reports: Will the Thyroid Imaging, Reporting, and Data System Improve Natural Language Processing Capture of Critical Thyroid Nodule Features?</title><source>MEDLINE</source><source>Elsevier ScienceDirect Journals Complete</source><creator>Chen, Kallie J. ; Dedhia, Priya H. ; Imbus, Joseph R. ; Schneider, David F.</creator><creatorcontrib>Chen, Kallie J. ; Dedhia, Priya H. ; Imbus, Joseph R. ; Schneider, David F.</creatorcontrib><description>Critical thyroid nodule features are contained in unstructured ultrasound (US) reports. The Thyroid Imaging, Reporting, and Data System (TI-RADS) uses five key features to risk stratify nodules and recommend appropriate intervention. This study aims to analyze the quality of US reporting and the potential benefit of Natural Language Processing (NLP) systems in efficiently capturing TI-RADS features from text reports.
This retrospective study used free-text thyroid US reports from an academic center (A) and community hospital (B). Physicians created “gold standard” annotations by manually extracting TI-RADS features and clinical recommendations from reports to determine how often they were included. Similar annotations were created using an automated NLP system and compared with the gold standard.
Two hundred eighty-two reports contained 409 nodules at least 1-cm in maximum diameter. The gold standard identified three nodules (0.7%) which contained enough information to calculate a complete TI-RADS score. Shape was described most often (92.7% of nodules), whereas margins were described least often (11%). A median number of two TI-RADS features are reported per nodule. The NLP system was significantly less accurate than the gold standard in capturing echogenicity (27.5%) and margins (58.9%). One hundred eight nodule reports (26.4%) included clinical management recommendations, which were included more often at site A than B (33.9 versus 17%, P < 0.05).
These results suggest a gap between current US reporting styles and those needed to implement TI-RADS and achieve NLP accuracy. Synoptic reporting should prompt more complete thyroid US reporting, improved recommendations for intervention, and better NLP performance.</description><identifier>ISSN: 0022-4804</identifier><identifier>EISSN: 1095-8673</identifier><identifier>DOI: 10.1016/j.jss.2020.07.015</identifier><identifier>PMID: 32799005</identifier><language>eng</language><publisher>United States: Elsevier Inc</publisher><subject>Academic Medical Centers - standards ; Academic Medical Centers - statistics & numerical data ; Data Systems ; Hospitals, Community - standards ; Hospitals, Community - statistics & numerical data ; Humans ; Image Processing, Computer-Assisted - methods ; Natural Language Processing ; Practice Guidelines as Topic ; Radiology - standards ; Retrospective Studies ; Societies, Medical - standards ; Thyroid ; Thyroid Gland - diagnostic imaging ; Thyroid Nodule - diagnosis ; TI-RADS ; Ultrasonography - standards ; Ultrasonography - statistics & numerical data ; Ultrasound</subject><ispartof>The Journal of surgical research, 2020-12, Vol.256, p.557-563</ispartof><rights>2020 Elsevier Inc.</rights><rights>Copyright © 2020 Elsevier Inc. All rights reserved.</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c326t-83b805a93ffe1a7b9d9fb5ad76bcdb9b7845a31bb0c51edcfe2ae9b024803beb3</citedby><cites>FETCH-LOGICAL-c326t-83b805a93ffe1a7b9d9fb5ad76bcdb9b7845a31bb0c51edcfe2ae9b024803beb3</cites><orcidid>0000-0002-1884-408X ; 0000-0002-2982-3583</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://dx.doi.org/10.1016/j.jss.2020.07.015$$EHTML$$P50$$Gelsevier$$H</linktohtml><link.rule.ids>314,780,784,3550,27924,27925,45995</link.rule.ids><backlink>$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/32799005$$D View this record in MEDLINE/PubMed$$Hfree_for_read</backlink></links><search><creatorcontrib>Chen, Kallie J.</creatorcontrib><creatorcontrib>Dedhia, Priya H.</creatorcontrib><creatorcontrib>Imbus, Joseph R.</creatorcontrib><creatorcontrib>Schneider, David F.</creatorcontrib><title>Thyroid Ultrasound Reports: Will the Thyroid Imaging, Reporting, and Data System Improve Natural Language Processing Capture of Critical Thyroid Nodule Features?</title><title>The Journal of surgical research</title><addtitle>J Surg Res</addtitle><description>Critical thyroid nodule features are contained in unstructured ultrasound (US) reports. The Thyroid Imaging, Reporting, and Data System (TI-RADS) uses five key features to risk stratify nodules and recommend appropriate intervention. This study aims to analyze the quality of US reporting and the potential benefit of Natural Language Processing (NLP) systems in efficiently capturing TI-RADS features from text reports.
This retrospective study used free-text thyroid US reports from an academic center (A) and community hospital (B). Physicians created “gold standard” annotations by manually extracting TI-RADS features and clinical recommendations from reports to determine how often they were included. Similar annotations were created using an automated NLP system and compared with the gold standard.
Two hundred eighty-two reports contained 409 nodules at least 1-cm in maximum diameter. The gold standard identified three nodules (0.7%) which contained enough information to calculate a complete TI-RADS score. Shape was described most often (92.7% of nodules), whereas margins were described least often (11%). A median number of two TI-RADS features are reported per nodule. The NLP system was significantly less accurate than the gold standard in capturing echogenicity (27.5%) and margins (58.9%). One hundred eight nodule reports (26.4%) included clinical management recommendations, which were included more often at site A than B (33.9 versus 17%, P < 0.05).
These results suggest a gap between current US reporting styles and those needed to implement TI-RADS and achieve NLP accuracy. Synoptic reporting should prompt more complete thyroid US reporting, improved recommendations for intervention, and better NLP performance.</description><subject>Academic Medical Centers - standards</subject><subject>Academic Medical Centers - statistics & numerical data</subject><subject>Data Systems</subject><subject>Hospitals, Community - standards</subject><subject>Hospitals, Community - statistics & numerical data</subject><subject>Humans</subject><subject>Image Processing, Computer-Assisted - methods</subject><subject>Natural Language Processing</subject><subject>Practice Guidelines as Topic</subject><subject>Radiology - standards</subject><subject>Retrospective Studies</subject><subject>Societies, Medical - standards</subject><subject>Thyroid</subject><subject>Thyroid Gland - diagnostic imaging</subject><subject>Thyroid Nodule - diagnosis</subject><subject>TI-RADS</subject><subject>Ultrasonography - standards</subject><subject>Ultrasonography - statistics & numerical data</subject><subject>Ultrasound</subject><issn>0022-4804</issn><issn>1095-8673</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2020</creationdate><recordtype>article</recordtype><sourceid>EIF</sourceid><recordid>eNp9kd1u0zAYhi0EYt3gAjhBPuSAZJ_jpEngYEJlY5OqMY1NHFr--dK5SuJiO5N6OdwpLm055Mi2_Lyvfx5C3jHIGbD5-Tpfh5AXUEAOdQ6sekFmDNoqa-Y1f0lmAEWRlQ2UJ-Q0hDWkdVvz1-SEF3XbAlQz8vvhaeudNfSxj14GN42G3uPG-Rg-0Z-272l8QnqEbga5suPq4wH5O5Up8VVGSX9sQ8QhMRvvnpHeyjh52dOlHFeTXCG9805jCClEF3KTNpG6ji68jVYn7njGrTNTj_QKd3kMF2_Iq072Ad8exjPyeHX5sLjOlt-_3Sy-LDPNi3nMGq4aqGTLuw6ZrFVr2k5V0tRzpY1qVd2UleRMKdAVQ6M7LCS2Cor0P1yh4mfkw743Xf_XhCGKwQaNfS9HdFMQRcnLuqoAmoSyPaq9C8FjJzbeDtJvBQOxMyPWIpkROzMCapHMpMz7Q_2kBjT_EkcVCfi8BzA98tmiF0FbHDUa61FHYZz9T_0fiSyibQ</recordid><startdate>202012</startdate><enddate>202012</enddate><creator>Chen, Kallie J.</creator><creator>Dedhia, Priya H.</creator><creator>Imbus, Joseph R.</creator><creator>Schneider, David F.</creator><general>Elsevier Inc</general><scope>CGR</scope><scope>CUY</scope><scope>CVF</scope><scope>ECM</scope><scope>EIF</scope><scope>NPM</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7X8</scope><orcidid>https://orcid.org/0000-0002-1884-408X</orcidid><orcidid>https://orcid.org/0000-0002-2982-3583</orcidid></search><sort><creationdate>202012</creationdate><title>Thyroid Ultrasound Reports: Will the Thyroid Imaging, Reporting, and Data System Improve Natural Language Processing Capture of Critical Thyroid Nodule Features?</title><author>Chen, Kallie J. ; Dedhia, Priya H. ; Imbus, Joseph R. ; Schneider, David F.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c326t-83b805a93ffe1a7b9d9fb5ad76bcdb9b7845a31bb0c51edcfe2ae9b024803beb3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2020</creationdate><topic>Academic Medical Centers - standards</topic><topic>Academic Medical Centers - statistics & numerical data</topic><topic>Data Systems</topic><topic>Hospitals, Community - standards</topic><topic>Hospitals, Community - statistics & numerical data</topic><topic>Humans</topic><topic>Image Processing, Computer-Assisted - methods</topic><topic>Natural Language Processing</topic><topic>Practice Guidelines as Topic</topic><topic>Radiology - standards</topic><topic>Retrospective Studies</topic><topic>Societies, Medical - standards</topic><topic>Thyroid</topic><topic>Thyroid Gland - diagnostic imaging</topic><topic>Thyroid Nodule - diagnosis</topic><topic>TI-RADS</topic><topic>Ultrasonography - standards</topic><topic>Ultrasonography - statistics & numerical data</topic><topic>Ultrasound</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Chen, Kallie J.</creatorcontrib><creatorcontrib>Dedhia, Priya H.</creatorcontrib><creatorcontrib>Imbus, Joseph R.</creatorcontrib><creatorcontrib>Schneider, David F.</creatorcontrib><collection>Medline</collection><collection>MEDLINE</collection><collection>MEDLINE (Ovid)</collection><collection>MEDLINE</collection><collection>MEDLINE</collection><collection>PubMed</collection><collection>CrossRef</collection><collection>MEDLINE - Academic</collection><jtitle>The Journal of surgical research</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>Chen, Kallie J.</au><au>Dedhia, Priya H.</au><au>Imbus, Joseph R.</au><au>Schneider, David F.</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Thyroid Ultrasound Reports: Will the Thyroid Imaging, Reporting, and Data System Improve Natural Language Processing Capture of Critical Thyroid Nodule Features?</atitle><jtitle>The Journal of surgical research</jtitle><addtitle>J Surg Res</addtitle><date>2020-12</date><risdate>2020</risdate><volume>256</volume><spage>557</spage><epage>563</epage><pages>557-563</pages><issn>0022-4804</issn><eissn>1095-8673</eissn><abstract>Critical thyroid nodule features are contained in unstructured ultrasound (US) reports. The Thyroid Imaging, Reporting, and Data System (TI-RADS) uses five key features to risk stratify nodules and recommend appropriate intervention. This study aims to analyze the quality of US reporting and the potential benefit of Natural Language Processing (NLP) systems in efficiently capturing TI-RADS features from text reports.
This retrospective study used free-text thyroid US reports from an academic center (A) and community hospital (B). Physicians created “gold standard” annotations by manually extracting TI-RADS features and clinical recommendations from reports to determine how often they were included. Similar annotations were created using an automated NLP system and compared with the gold standard.
Two hundred eighty-two reports contained 409 nodules at least 1-cm in maximum diameter. The gold standard identified three nodules (0.7%) which contained enough information to calculate a complete TI-RADS score. Shape was described most often (92.7% of nodules), whereas margins were described least often (11%). A median number of two TI-RADS features are reported per nodule. The NLP system was significantly less accurate than the gold standard in capturing echogenicity (27.5%) and margins (58.9%). One hundred eight nodule reports (26.4%) included clinical management recommendations, which were included more often at site A than B (33.9 versus 17%, P < 0.05).
These results suggest a gap between current US reporting styles and those needed to implement TI-RADS and achieve NLP accuracy. Synoptic reporting should prompt more complete thyroid US reporting, improved recommendations for intervention, and better NLP performance.</abstract><cop>United States</cop><pub>Elsevier Inc</pub><pmid>32799005</pmid><doi>10.1016/j.jss.2020.07.015</doi><tpages>7</tpages><orcidid>https://orcid.org/0000-0002-1884-408X</orcidid><orcidid>https://orcid.org/0000-0002-2982-3583</orcidid><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext |
| identifier | ISSN: 0022-4804 |
| ispartof | The Journal of surgical research, 2020-12, Vol.256, p.557-563 |
| issn | 0022-4804 1095-8673 |
| language | eng |
| recordid | cdi_proquest_miscellaneous_2434755008 |
| source | MEDLINE; Elsevier ScienceDirect Journals Complete |
| subjects | Academic Medical Centers - standards Academic Medical Centers - statistics & numerical data Data Systems Hospitals, Community - standards Hospitals, Community - statistics & numerical data Humans Image Processing, Computer-Assisted - methods Natural Language Processing Practice Guidelines as Topic Radiology - standards Retrospective Studies Societies, Medical - standards Thyroid Thyroid Gland - diagnostic imaging Thyroid Nodule - diagnosis TI-RADS Ultrasonography - standards Ultrasonography - statistics & numerical data Ultrasound |
| title | Thyroid Ultrasound Reports: Will the Thyroid Imaging, Reporting, and Data System Improve Natural Language Processing Capture of Critical Thyroid Nodule Features? |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-04T18%3A44%3A13IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_cross&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Thyroid%20Ultrasound%20Reports:%20Will%20the%20Thyroid%20Imaging,%20Reporting,%20and%20Data%20System%20Improve%20Natural%20Language%20Processing%20Capture%20of%20Critical%20Thyroid%20Nodule%20Features?&rft.jtitle=The%20Journal%20of%20surgical%20research&rft.au=Chen,%20Kallie%20J.&rft.date=2020-12&rft.volume=256&rft.spage=557&rft.epage=563&rft.pages=557-563&rft.issn=0022-4804&rft.eissn=1095-8673&rft_id=info:doi/10.1016/j.jss.2020.07.015&rft_dat=%3Cproquest_cross%3E2434755008%3C/proquest_cross%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2434755008&rft_id=info:pmid/32799005&rft_els_id=S0022480420304728&rfr_iscdi=true |