[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-331243-105":59,"doc-detail-331243-en":131},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":124,"head_meta":126,"extra_data":128,"updated_unix":130},105,"en","use-of-natural-language-processing-to-infer-sites-of-metastatic-disease-from-radiology-reports-at-scale","Use of Natural Language Processing to Infer Sites of Metastatic Disease From Radiology Reports at Scale","","To evaluate natural language processing methods for inferring metastatic sites from radiology reports. A dataset of 4,522 computed tomography (CT) reports from 550 patients across 14 cancer types was used to fine-tune four clinical large language models for multilabel classification. An information extraction system based on named entity recognition, assertion status detection, and relation extraction was also built for comparison and assessed with F1 scores on test and external validation sets. The IE system achieved the highest performance and supported cohort-level analyses that differentiate de novo versus recurrent metastatic patterns.",{"@graph":69,"@context":123},[70,84,106],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/use-of-natural-language-processing-to-infer-sites-of-metastatic-disease-from-radiology-reports-at-scale/331243/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/use-of-natural-language-processing-to-infer-sites-of-metastatic-disease-from-radiology-reports-at-scale/331243.png","ImageObject",300,407,{"name":92,"@type":93},"Asher","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-11-08","2026-09-21",true,{"@type":102,"interactionType":103,"userInteractionCount":105},"InteractionCounter",{"@type":104},"ViewAction",12,{"@type":107,"mainEntity":108},"FAQPage",[109,115,119],{"name":110,"@type":111,"acceptedAnswer":112},"What is the main goal of this study?","Question",{"text":113,"@type":114},"The study evaluates natural language processing methods to infer metastatic sites from unstructured radiology reports across multiple primary cancers.","Answer",{"name":116,"@type":111,"acceptedAnswer":117},"How were the models evaluated?",{"text":118,"@type":114},"Model performance was measured using F1 scores on a test set and three external validation sets.",{"name":120,"@type":111,"acceptedAnswer":121},"What approach performed best for metastatic-site inference?",{"text":122,"@type":114},"The information extraction (IE) system performed best, achieving the highest F1 score on the test set compared with the clinical large language models.","https://schema.org",{"og:url":83,"og:type":125,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":127,"canonical":83},"index,follow",{"doc_id":129,"site_id":62},331243,1791163686,{"code":4,"msg":5,"data":132},{"doc_id":129,"user_id":133,"nickname":92,"user_avatar":134,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":135,"file_id":136,"file_url":137,"file_type":138,"file_size":139,"view_count":105,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":52,"language":140,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":67,"update_tm":144,"read_time":145,"preview_page_count":4},687197207639,"https://ap-avatar.wpscdn.com/davatar_a8503ba1806abce46bf441b54a3ca4cd","Original Reports | Artificial Intelligence  \nUse of Natural Language Processing to Infer Sites of Metastatic Disease From Radiology Reports at Scale  \nSee Boon Tay, MBBS1,2 ; Guat Hwa Low, BS1,3 ; Gillian Jing En Wong, MBBS2; Han Jieh Tey, MSc1,3; Fun Loon Leong, BS1,3  ; Constance Li, PhD3 ; Melvin Lee Kiang Chua, MBBS, PhD3,4,5 ; Daniel Shao Weng Tan, MBBS, PhD1,4,6 ; Choon Hua Thng, MBBS4,7; Iain Bee Huat Tan, MBBS, PhD1,3,4; and Ryan Shea Ying Cong Tan, MBBS, MTec1,3,4,8   \nDOI [https://doi.org/10.1200/CCI.23.00122](https://doi.org/10.1200/CCI.23.00122)  \n\n| ABSTRACT |  |\n| --- | --- |\n| PURPOSE | To evaluate natural language processing (NLP) methods to infer metastatic sites from radiology reports. |\n| METHODS | A set of 4,522 computed tomography (CT) reports of 550 patients with 14 types of cancer was used to fine-tune four clinical large language models (LLMs) for multilabel classification of metastatic sites. We also developed an NLP information extraction (IE) system (on the basis of named entity recognition, assertion status detection, and relation extraction) for comparison. Model performances were measured by F1 scores on test and three external validation sets. The best model was used to facilitate analysis of metastatic frequencies in a cohort study of 6,555 patients with 53,838 CT reports. |\n| RESULTS | The RadBERT, BioBERT, GatorTron-base, and GatorTron-medium LLMs achieved F1 scores of 0.84, 0.87, 0.89, and 0.91, respectively, on the test set. The IE system performedbest, achieving anF1 score of 0.93. F1 scores ofthe IE system by individual cancer type ranged from 0.89 to 0.96. The IE system attained F1 scores of 0.89, 0.83, and 0.81, respectively, on external validation sets including additional cancer types, positron emission tomography-CT ,and magnetic resonance imaging scans, respectively. In our cohort study, we found that for colorectal cancer, liver-only metastases were higher in de novo stage IV versus recurrent patients (29.7% v 12.2%; P \u003C .001). Conversely, lung-only metastases were more frequent in recurrent versus de novo stage IV patients (17.2% v 7.3%; P \u003C .001) . |\n| CONCLUSION | We developed an IE system that accurately infers metastatic sites in multiple primary cancers from radiology reports. It has explainable methods and performs better than some clinical LLMs. The inferred metastatic phenotypes could enhance cancer research databases and clinical trial matching, and identify potential patients for oligometastatic interventions. |\n\nACCOMPANYING CONTENT  \n Data Supplement  \nAccepted April 1, 2024  \nPublished May 24, 2024  \nJCO Clin Cancer Inform 8:e2300122  \n© 2024 by American Society of Clinical Oncology  \nCreative Commons Attribution Non-Commercial No Derivatives 4.0 License  \nINTRODUCTION  \nPattern of metastatic spread is an important diagnostic and prognostic factor in advanced-stage cancer.1 ,2 Different primary cancers display distinct metastatic organotropism, which may lend insight into disease biology.3 For instance, bone-only tropism is seen more frequently in breast and prostate cancers, and often carries a better prognosis than those with visceral metastases.4-6 Metastatic phenotype and its evolution over time guide clinical decision making and treatment selection. For example, surgical and radiotherapy treatments for oligometastatic disease can be discussed while clinical trials may have specific criteria such as excluding patients with active brain metastases.7,8 Thus, it is important for oncology electronic health records tobe able to accurately capture this information for every patient.  \nComputed tomography (CT) is a common imaging modality for identifying sites of metastatic disease.9 Data extraction from CT radiology reports using natural language processing (NLP) has been explored to gain insights from this rich source of unstructured clinical data.10-12 Earlier work to identify sites of metastases used conventional machine learning methods (eg, XGBoost classifiers) and early","cbCaij8X2en74Cjl","https://ap.wps.com/l/cbCaij8X2en74Cjl","pdf",548029,"English","# Abstract\n## Purpose\n## Methods\n## Results\n## Conclusion\n# Introduction\n# Context\n## Key Objective\n## Knowledge Generated\n## Relevance","[{\"question\":\"What is the main goal of this study?\",\"answer\":\"The study evaluates natural language processing methods to infer metastatic sites from unstructured radiology reports across multiple primary cancers.\"},{\"question\":\"How were the models evaluated?\",\"answer\":\"Model performance was measured using F1 scores on a test set and three external validation sets.\"},{\"question\":\"What approach performed best for metastatic-site inference?\",\"answer\":\"The information extraction (IE) system performed best, achieving the highest F1 score on the test set compared with the clinical large language models.\"}]","Use of Natural Language Processing to Infer Sites of Metastatic Disease From Radiology Reports at Scale | PDF",1789992374,25]