Mô hình khai thác khái niệm và quan hệ trong văn bản lâm sàng

ĐẠI HỌC QUỐC GIA TP HỒ CHÍ MINH TRƯỜNG ĐẠI HỌC KHOA HỌC TỰ NHIÊN HUỲNH HỮU NGHĨA MƠ HÌNH KHAI THÁC KHÁI NIỆM VÀ QUAN HỆ TRONG VĂN BẢN LÂM SÀNG LUẬN ÁN TIẾN SĨ CÔNG NGHỆ THÔNG TIN Tp Hồ Chí Minh – Năm 2017 ĐẠI HỌC QUỐC GIA TP HỒ CHÍ MINH TRƯỜNG ĐẠI HỌC KHOA HỌC TỰ NHIÊN HUỲNH HỮU NGHĨA MƠ HÌNH KHAI THÁC KHÁI NIỆM VÀ QUAN HỆ TRONG VĂN BẢN LÂM SÀNG Chuyên ngành : HỆ THỐNG THÔNG TIN Mã số : 62.48.05.01 Phản biện : PGS.TS Lê Anh Cường Phản biện : TS Võ Thị Ngọc Châu Phản biện : TS Nguyễn Tuấn Đăng Phản biện độc lập : TS Võ Thị Ngọc Châu Phản biện độc lập : TS Nguyễn Tuấn Đăng NGƯỜI HƯỚNG DẪN KHOA HỌC: PGS TS HỒ BẢO QUỐC TS NGUYỄN AN TẾ Tp Hồ Chí Minh – Năm 2017 LỜI CAM ĐOAN Tôi xin cam đoan luận án cơng trình nghiên cứu khoa học tơi với hướng dẫn PGS TS Hồ Bảo Quốc TS Nguyễn An Tế Các kết luận án trung thực chưa công bố cơng trình khác Tp Hồ Chí Minh, ngày 09 tháng 11 năm 2017 Người thực Huỳnh Hữu Nghĩa LỜI CẢM ƠN Tôi chân thành bày tỏ lòng biết ơn sâu sắc đến PGS TS Hồ Bảo Quốc TS Nguyễn An Tế Quý Thầy tận tình hướng dẫn tơi suốt q trình làm luận án Tôi Quý Thầy cung cấp tài liệu, bảo tận tình phong cách làm việc, phương pháp nghiên cứu khoa học tạo điều kiện thuận lợi cho học tập nghiên cứu Xin bày tỏ lịng biết ơn đến Q Thầy Cơ Khoa Công nghệ thông tin, Trường Đại học Khoa học Tự nhiên, Đại học Quốc gia Tp Hồ Chí Minh Quý Thầy Cô trang bị cho kiến thức quý báu để thực luận án Cảm ơn gia đình bạn bè đồng nghiệp quan tâm giúp đỡ tơi q trình học tập, nghiên cứu Tôi cảm ơn tác giả cơng trình nghiên cứu mà luận án trích dẫn tham khảo tổ chức diễn đàn khoa học mà tơi tham gia q trình nghiên cứu Tp Hồ Chí Minh, ngày 09 tháng 11 năm 2017 Huỳnh Hữu Nghĩa MỤC LỤC DANH MỤC TỪ VIẾT TẮT DANH MỤC HÌNH DANH MỤC BẢNG BIỂU MỞ ĐẦU Dẫn nhập Lý thực mục tiêu đề tài Nội dung nghiên cứu luận án 10 Phạm vi nghiên cứu 11 Phương pháp nghiên cứu 11 Những đóng góp luận án 11 Bố cục luận án 12 Chương 14 GIỚI THIỆU 14 1.1 Dữ liệu y tế 14 1.2 Khai thác văn tổng quát 15 1.2.1 Rút trích thơng tin 16 1.2.1.1 Nhận diện thực thể định danh .20 1.2.1.2 Rút trích mối quan hệ 24 1.2.2 Các chuẩn đánh giá hệ thống rút trích thơng tin 25 1.3 Khai thác văn y tế 27 1.3.1 Xử lý ngôn ngữ tự nhiên lĩnh vực y tế 27 1.3.1.1 Xử lý mức hình thái 28 1.3.1.2 Xử lý mức cú pháp 30 1.3.1.3 Xử lý mức ngữ nghĩa 30 1.3.1.4 Xử lý mức ngữ dụng .31 1.3.2 Rút trích khái niệm 32 1.3.2.1 Các thách thức .33 1.3.2.2 Các phương pháp rút trích 35 1.3.3 Rút trích mối quan hệ 38 1.3.3.1 Những thách thức rút trích mối quan hệ .38 1.3.3.2 Một số đặc điểm xác định mối quan hệ 40 1.3.3.3 Các phương pháp rút trích mối quan hệ 42 1.3.4 Tóm tắt tài liệu y tế 45 1.3.5 Hỏi – Đáp thông tin y tế .49 1.3.5.1 Hỏi – Đáp lâm sàng 49 1.3.5.2 Trả lời câu hỏi y sinh học .50 1.3.6 Các hệ thống rút trích thông tin y tế 51 1.3.7 Các kiện khai thác văn y tế 51 1.3.8 Các nguồn tài nguyên phục vụ khai thác văn y tế 53 1.3.8.1 Kho ngữ liệu 53 1.3.8.2 Dữ liệu gán nhãn 55 1.3.8.3 Các nguồn tri thức .56 1.3.8.4 Các công cụ hỗ trợ 58 1.4 Kết luận 59 Chương 61 RÚT TRÍCH KHÁI NIỆM Y TẾ 61 2.1 Giới thiệu 61 2.2 Những đề xuất liên quan 63 2.2.1 Bộ nhãn phân lớp Token 64 2.2.2 Tập đặc trưng phân lớp Token .66 2.2.3 Hệ thống rút trích chuẩn hóa khái niệm 67 2.2.3.1 Xử lý ngôn ngữ tự nhiên .68 2.2.3.2 Trích khái niệm huấn luyện 73 2.2.3.3 Gán nhãn token .74 2.2.3.4 Xây dựng mơ hình 74 2.2.3.5 Dự đoán nhãn Token .74 2.2.3.6 Trích xuất khái niệm .75 2.2.3.7 Chuẩn hóa khái niệm 79 2.3 Áp dụng đề xuất 80 2.4 Đánh giá hiệu đề xuất 82 2.4.1 Phương pháp đánh giá 82 2.4.2 Dữ liệu thực nghiệm 83 2.4.3 Kết đề xuất 84 Rút trích khái niệm 85 Chuẩn hóa khái niệm 89 2.5 Tổng kết 92 Chương 93 RÚT TRÍCH MỐI QUAN HỆ GIỮA CÁC KHÁI NIỆM 93 3.1 Giới thiệu 93 3.2 Các toán rút trích mối quan hệ lĩnh vực y tế 95 3.2.1 Bài toán mối quan hệ nhị phân 95 3.2.2 Bài toán quan hệ đồng tham chiếu 97 3.2.3 Bài toán quan hệ thời gian .98 3.2.4 Bài toán điền mẫu 100 3.2.5 Bài tốn rút trích kiện 100 3.3 Đề xuất liên quan đến toán điền mẫu 101 3.3.1 Tập luật xác định mối quan hệ .102 3.3.2 Tập đặc trưng phân lớp mối quan hệ thời gian 109 3.3.3 Hệ thống điền mẫu .113 3.3.3.1 Tiền xử lý 124 3.3.3.2 Xử lý ngôn ngữ tự nhiên .125 3.3.3.3 Rút trích dấu hiệu (trigger) 126 3.3.3.4 Rút trích luật .127 3.3.3.5 Rút trích đặc trưng 128 3.3.3.6 Xác định giá trị thuộc tính 128 3.3.4 Đánh giá đề xuất .135 3.3.4.1 Dữ liệu thực nghiệm 135 3.3.4.2 Phương pháp đánh giá 136 3.3.4.3 Kết 136 3.4 Đề xuất liên quan đến toán rút trích kiện y sinh 140 3.4.1 Tập đặc trưng phân lớp kiện 141 3.4.2 Tập luật xác định tham số kiện y sinh .142 3.4.3 Hệ thống rút trích kiện 144 3.4.3.1 Xử lý ngôn ngữ tự nhiên .146 3.4.3.2 Rút trích luật .148 3.4.3.3 Trích đặc trưng 148 3.4.3.4 Phân lớp kiện 148 3.4.3.5 Xác định tham số kiện 148 3.4.4 Đánh giá hiệu đề xuất .148 3.5 Tổng kết 151 Chương 152 KIẾN TRÚC HỆ THỐNG 152 KHAI THÁC KHÁI NIỆM VÀ MỐI QUAN HỆ 152 4.1 Giới thiệu 152 4.2 Mẫu hệ thống khai thác thông tin y tế dựa khái niệm 154 4.3.1 Tài liệu y tế 156 4.3.2 Xử lý ngôn ngữ tự nhiên 156 4.3.3 Rút trích khái niệm 156 4.3.4 Rút trích mối quan hệ 157 4.3.5 Nguồn tài nguyên 157 4.3.6 Kho tri thức 157 4.3.7 Ứng dụng .158 4.3 Hệ thống thực nghiệm 158 4.4.1 Tài liệu lâm sàng 160 4.4.2 Xử lý ngôn ngữ tự nhiên 160 4.4.3 Nguồn tài nguyên 161 4.4.4 Rút trích khái niệm 161 4.4.5 Rút trích mối quan hệ 161 4.4.6 Kho tri thức 162 4.4.7 Lập mục tài liệu .162 4.4.8 Ứng dụng .164 4.4 Định hướng khung kiến trúc chung khai thác tài liệu y tế 167 4.5.1 Quan điểm GATE 168 4.5.1.1 Khung kiến trúc chung 169 4.5.1.2 Các đối tượng xử lý tài liệu 170 4.5.2 Định hướng khung kiến trúc chung khai thác tài liệu y tế 172 4.5.3 Khung kiến trúc đề xuất (MedTMF) GATE 174 4.5 Tổng kết 174 Chương 175 KẾT LUẬN VÀ HƯỚNG PHÁT TRIỂN 175 5.1 Kết luận 175 5.2 Hướng phát triển 178 DANH MỤC CÔNG TRÌNH KHOA HỌC 180 TÀI LIỆU THAM KHẢO 182 Phụ lục A: 193 DANH MỤC TỪ VIẾT TẮT ACE Automatic Content Extraction CRF Conditional Random Fields EHR Electronic Health Record FSM Finite State Machines I2B2 Informatics for Integrating Biology and the Bedside MeSH Medical Subject Headings MUC Message Understanding Conference NER Named Entity Recognition NLM National Library of Medicine NLP Natural Language Processing SNOMED-CT Systematized Nomenclature of Medicine - Clinical Terms SVM Support Vector Machine UIMA Unstructured Information Management Architecture UMLS Unified Medical Language System XML eXtensible Markup Language [17] Erik F Tjong Kim Sang and Fien De Meulder Introduction to the CoNLL-2003 shared task: Language-independent named entity recognition In Proceedings of the 7th Conference on Natural Language Learning, pages 142–147, 2003 [18] Facult, Hans P.; Paulussen, Hans; Martin, Willy: Dilemma-2: A Lemmatizer-Tagger For Medical Abstracts In: Third Conference on Applied Language Processing, 1992 [19] Fei Wu and Daniel S Weld Open information extraction using Wikipedia In Proceedings of the 48th Annual Meeting of the Association for Computational Linguistics, pages 118–127, 2010 [20] Fiszman M., Demner-Fushman D., Kilicoglu H., and Rindflesch T C., Automatic summarization of MEDLINE citations for evidence-based medical treatment: A topic-oriented evaluation Journal of Biomedical Informatics, 42(5): pages 801–813, 2009 [21] Fiszman M., Demner-Fushman D., Kilicoglu H., and Rindflesch T C., Automatic summarization of MEDLINE citations for evidence-based medical treatment: A topic-oriented evaluation Journal of Biomedical Informatics, 42(5):801–813, 2009 [22] Friedman, C., ed Sublanguage - Zellig Harris Memorial A special issue of the Journal of Biomedical Informatics devoted to Zellig Harris‟s theory on sublanguage processing J Biomed Inform 35: pages 213-277, 2002 [23] Gerald DeJong Prediction and substantiation: A new approach to natural language processing Cognitive Science, 3: pages 251–173, 1979 [24] Grinberg D., Lafferty J and Sleator D A robust parsing algorithm for link grammars Proceedings of the Fourth International Workshop on Parsing Technologies, Prague and Karlovy Vary, September 1995, pages 111-125 Also issued as technical report CMU-CS-95-125, Department of Computer Science, Carnegie Mellon University, 1995 184 [25] Gurulingappa, Harsha; Klinger, Roman; Hofmann-Apitius, Martin; Fluck, Juliane: An Empirical Evaluation of Resources for the Identification of Diseases and Adverse Effects in Biomedical Literature In: 2nd Workshop on Building and evaluating resources for biomedical text mining (7th edition of the Language Resources and Evaluation Conference), 2010 [26] Hamon, Thierry; Grabar, Natalia: Linguistic approach for identification of medication names and related information in clinical narratives In: J Am Med Inform Assoc 17 (2010), No 5, pages 549–554 – URL http://dx.doi.org/10.1136/ jamia.2010.004036 [27] Hanisch, Daniel; Fundel, Katrin; Mevissen, Heinz-Theodor; Zimmer, Ralf; Fluck, Juliane: ProMiner: rule-based protein and gene entity recognition In: BMC Bioinformatics Suppl (2005), pages S14 – URL http://dx.doi.org/10.1186/14712105-6-S1-S14 [28] Hawizy, Lezan; Jessop, David M.; Adams, Nico; Murray-Rust, Peter: ChemicalTagger: A tool for semantic text-mining in chemistry In: J Cheminform (2011), No 1, pages 17 – URL http://dx.doi.org/10.1186/1758-2946-3-17 [29] Hettne, Kristina M.; Stierum, Rob H.; Schuemie, Martijn J.; Hendriksen, Peter J M.; Schijvenaars, Bob J A.; Mulligen, Erik M v.; Kleinjans, Jos; Kors, Jan A.: A dictionary to identify small molecules and drugs in free text In:Bioinformatics 25 (2009), Nov, No 22, pages 2983–2991 – URL http://dx.doi.org/10.1093/ bioinformatics/btp535 [30] Huang X., Lin J., and Demner-Fushman D., Evaluation of PICO as a knowledge representation for clinical questions In AMIA Annual Symposium Proceedings, pages 359–363, 2006 [31] Jiang, Jing; Zhai, Cheng X.: An empirical study of tokenization strategies for biomedical information retrieval In: Information Retrieval 10 (2007), pages 341– 363 185 [32] Kim S., Yoon J., and Yang J Kernel approaches for genic interaction extraction Bioinformatics, 24(1): pages 118–126, 2008 [33] Kim J.-D., T., Pyysalo S., Kano Y., and Tsujii J Overview of BioNLP‟09 shared Ohta task on event extraction In Proceedings of the Workshop on Current Trends in Biomedical Natural Language Processing: Shared Task, pages 1–9, 2009 [34] Kim J.-D., Ohta T., Nguyen N., Pyysalo S., Bossy R., and Tsujii J Overview of BioNLP shared task 2011 In Proceedings of the BioNLP Shared Task 2011 Workshop, pages 1–6, 2011 [35] Klinger, Roman; Friedrich, Christoph M.; Mevissen, Heinz T.; Fluck, Juliane; Hofmann-Apitius, Martin; Furlong, Laura I.; Sanz, Ferran: Identifying gene-specific variations in biomedical text In: J Bioinform Comput Biol (2007), Dec, No 6, pages 1277–1296 [36] Klinger, Roman; Kolárik, Corinna; Fluck, Juliane; Hofmann-Apitius, Martin; Friedrich, Christoph M.: Detection of IUPAC and IUPAC-like chemical names In: Bioinformatics 24 (2008), Jul, No 13, pages i268–i276 – URLhttp://dx.doi.org/10.1093/bioinformatics/btn181 [37] Kontos J., Malagardi I., and Peros J., Question answering and rhetoric analysis of biomedical texts in the AROMA system In Proceedings of the 7th Hellenic Europeoan Conference on Computer Mathematics and its Applications, 2005 [38] Krauthammer, Michael; Nenadic, Goran: Term identification in the biomedical literature In: J Biomed Inform 37 (2004), Dec, No 6, pages 512–526 – URL http://dx.doi.org/10.1016/j.jbi.2004.08.004 [39] Lance A Ramshaw and Mitch P Marcus Text chunking using transformation-based learning In Proceedings of the 3rd Workship on Very Large Corpora, pages 82–94, 1995 186 [40] Leaman, Robert; Gonzalez, Graciela: BANNER: an executable survey of advances in biomedical named entity recognition In: Pac Symp Biocomput (2008), pages 652– 663 [41] Lin C.-Y., ROUGE: A package for automatic evaluation of summaries In Proceedings of the Workshop on Text Summarization Branches Out, 2004 [42] Lin C.-Y., Cao G., Gao J., and Nie J.-Y An information-theoretic approach to automatic evaluation of summaries In Proceedings of the Human Language Technology Conference of the North American Chapter of the Association of Computational Linguistics, pages 463–470, 2006 [43] Lin R T K., Liang-Te Chiu J., Dai H.-J., Day M.-Y., Tsai R T.-H., and Hsu W.-L., Biological question answering with syntactic and semantic feature matching and an improved mean reciprocal ranking measurement In Proceedings of the 2008 IEEE International Conference on Information Reuse and Integration, pages 184–189, 2008 [44] Ling X., Jiang J., He X., Mei Q., Zhai C., and Schatz B Generating gene summaries from biomedical literature: A study of semi-structured summarization Information Processing & Management, 43(6): pages 1777–1791, 2007 [45] Liu, Hongfang, Yves A Lussierb, and Carol Friedman (2001), “Disambiguating ambiguous biomedical terms in biomedical narrative text: An unsupervised method." Journal of Biomedical Informatics, 34, pages 249-261 [46] Mahbub Chowdhury, Faisal; Lavelli, Alberto: Disease Mention Recognition with Specific Features In: Proceedings of the 2010 Workshop on Biomedical Natural Language Processing (BioNLP), ACL 2010, pages 83–90 [47] Marie-Catherine de Marneffe and Christopher D Manning Stanford typed dependencies manual, September 2008, revised for the Stanford Parser v 3.3 in December 2013 187 [48] Marie-Catherine de Marneffe and Christopher D Manning Stanford typed dependencies manual September 2008 Revised for the Stanford Parser v 3.5.2 in April 2015 [49] Melinda Katona and Richárd Farkas SZTE-NLP: Clinical Text Analysis with Named Entity Recognition Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval 2014), pages 615–618, Dublin, Ireland, August 2324, 2014 [50] Michele Banko, Michael J Cafarella, Stephen Soderland, Matthew Broadhead, and Oren Etzioni Open information extraction from the Web In Proceedings of the 20th International Joint Conference on Artificial Intelligence, pages 2670–2676, 2007 [51] Miwa M., Sætre R., Miyao Y., and Tsujii J Protein-protein interaction extraction by leveraging multiple kernels and parsers International Journal of Medical Informatics, 78(12):e39–e46, 2009 [52] Morante, R.; Daelemans, W.: Learning the scope of hedge cues in biomedical texts In: Workshop on BioNLP, 2009, pages 28–36 [53] Niu Y., Zhu X., and Hirst G., Using outcome polarity in sentence extraction for medical question-answering In AMIA Annual Symposium Proceedings, pages 599– 603, 2006 [54] Noémie Elhadad, Sameer Pradhan, Sharon Lipsky Gorman, Suresh Manandhar, Wendy Chapman, Guergana Savova SemEval-2015 Task 14: Analysis of Clinical Text Proceedings of the 9th International Workshop on Semantic Evaluation (SemEval 2015), 2015 Association for Computational Linguistics, Denver, Colorado, June 4-5, 2015, pages 303–310 [55] Özlem Uzuner, Ira Goldstein, Yuan Luo, Isaac Kohane Identifying Patient Smoking Status from Medical Discharge Records Journal of the American Medical Informatics Association Volume 15 Number Jan / Feb 2008, pages 14-24 188 [56] Özlem Uzuner Recognizing obesity and comorbidities in sparse data Journal of the American Medical Informatics Association: JAMIA, 16(4), 2009, pages 561–570 [57] Özlem Uzuner, Imre Solti, Eithon Cadag Extracting medication information from clinical text J Am Med Inform Assoc 2010;17: pages 514-518 [58] Pustejovsky J, Hanks P, Sauri R, et al The timebank corpus Proceeding of Corpus Linguistics, Lancaster University, UK, 2003, pages 647–56 [59] Pyysalo S., Airola A., Heimonen J., Bjorne J., Ginter F., and Salakoski T., Comparative analysis of five protein-protein interaction corpora BMC Bioinformatics, 9(Suppl 3):S6, 2008 [60] Quint, Julien: A formalism for universal segmentation of text In: Proceeding COLING ‟00 Proceedings of the 18th conference on Computational linguistics, 2000 [61] Ramanan S V., Shereen Broido, and Senthil Nathan P Performance of a multi-class biomedical tagger on clinical records Online Working Notes of the CLEF 2013 Evaluation Labs and Workshop [62] Danielle L Mowery, Sumithra Velupillai, Brett R South, Lee Christensen, David Martinez, Liadh Kelly, Lorraine Goeuriot, Noemie Elhadad, Sameer Pradhan, Guergana Savova, and Wendy W Chapman Task 2: ShARe/CLEF eHealth Evaluation Lab 2014 Working Notes for CLEF 2014 Conference, pages 31 – 42, Sheffield, UK, September 15-18, 2014 ISSN 1613-0073, Vol-1180 [63] Richardson W S., Wilson M C., Nishikawa J., and Hayward R S., The well-built clinical question: A key to evidence-based decisions ACP Journal Club, 123(3): pages A12–A13, 1995 [64] Rinaldi F., Dowdall J., Schneider G., and Persidis A., Answering questions in the genomics domain In Proceedings of the ACL 2004 Workshop on Question Answering in Restricted Domains, 2005 189 [65] Rinaldi F., Kaljurand K., and Saetre R Terminological resources for text mining over biomedical scientific literature Artificial Intelligence in Medicine, 52(2): pages 107–114, 2011 [66] Rindflesch T C., Kilicoglu H., Fiszman M., Rosemblat G., and Shin D., Semantic MEDLINE: An advanced information management application for biomedicine Information Services & Use, 31: pages 15–21, 2011 [67] Rink B., Harabagiu S., and Roberts K Automatic extraction of relations between medical concepts in clinical texts Journal of the American Medical Informatics Association, 18(5): pages 594–600, 2011 [68] Roberts A., Gaizauskas R., and Hepple M Extracting clinical relationships from patient narratives In Proceedings of the Workshop on Current Trends in Biomedical Natural Language Processing, pages 10–18, 2008 [69] Ruch P., Boyer C., Chichester C., Tbahriti I., Geissbăuhler A., Fabry P., Gobeill J., Pillet V., Rebholz-Schuhmann D., Lovis C., and Veuthey A.-L Using argumentation to extract key sentences from biomedical abstracts International Journal of Medical Informatics, 76(2-3): pages 195–200, 2007 [70] Shatkay H., Wilbur J W., and Rzhetsky A., Annotation guidelines, 2005 http://www.ncbi.nlm.nih.gov/CBBresearch/Wilbur/AnnotationGuidelines.pdf [71] Smalheiser N., Zhou W., and Torvik V., Anne O‟Tate: A tool to support user-driven summarization, drill-down and browsing of PubMed search results Journal of Biomedical Discovery and Collaboration, 3(1):2, 2008 [72] Smith, L.; Rindflesch, T.; Wilbur,W J.: MedPost: a part-of-speech tagger for bioMedical text In: Bioinformatics 20 (2004), Sep, No 14, pages 2320–2321 – URL http://dx.doi.org/10.1093/bioinformatics/bth227 [73] Strötgen J, Gertz M HeidelTime High quality rule-based extraction and normalization of temporal expressions Proceedings of the 5th International 190 Workshop on Semantic Evaluation Los Angeles, California: Association for Computational Linguistics 2010 [74] Takahashi K., Koike A., and Takagi T., Question answering system in biomedical domain In Proceedings of the 15th International Conference on Genome Informatics, pages 161–162, 2004 [75] Tsuruoka, Yoshimasa; Tateishi, Yuka; Kim, Jin-Dong; Ohta, Tomoko; McNaught, John; Ananiadou, Sophia; Tsujii, Junichi: Developing a Robust Part-of-Speech Tagger for Biomedical Text In: Lecture Notes in Computer Science 3746 (2005), pages 382–392 [76] Valter Crescenzi, Giansalvatore Mecca, and Paolo Merialdo Road-Runner: Towards automatic data extraction from large Web sites In Proceedings of the 27th International Conference on Very Large Data Bases, pages 109–118, 2001 [77] Verspoora, K.; Roeder, C.; Johnson, H.; Cohen, K.; Baumgartner, W.; Hunter, L.: Information Extraction of Normalized Protein Interaction Pairs Utilizing Linguistic and Semantic Cues In: Proceedings of the BioCreative II Workshop 2009 on Digital Annotations, 2009 [78] Wattarujeekrit T., Shah P., and Collier N PASBio: Predicateargument structures for event extraction in molecular biology BMC Bioinformatics, 5(1): pages 155, 2004 [79] Wilbur J W., Rzhetsky A., and Shatkay H., New directions in biomedical text annotation: Definitions, guidelines and corpus construction BMC Bioinformatics, 7:356, 2006 [80] Yang J., Cohen A M., and Hersh W Automatic summarization of mouse gene information by clustering and sentence extraction from MEDLINE abstracts In AMIA Annual Symposium Proceedings, pages 831–835, 2007 [81] Yaoyun Zhang, Jingqi Wang, Buzhou Tang, Yonghui Wu, Min Jiang, Yukun Chen, and Hua Xu UTH_CCB: A Report for SemEval 2014 – Task Analysis of Clinical 191 Text Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval 2014), pages 802–806, Dublin, Ireland, August 23-24, 2014 [82] Yee Seng Chan and Dan Roth Exploiting background knowledge for relation extraction In Proceedings of the 23rd International Conference on Computational Linguistics, pages 152–160, 2010 [83] Yoo I., Hu X., and Song I.-Y., A coherent graph-based semantic clustering and summarization approach for biomedical literature and a new summarization evaluation method BMC Bioinformatics, 8(Suppl 9):S4, 2007 [84] Yu H., Lee M., Kaufman D., Ely J., Osheroff J A., Hripcsak G., and Cimino J., Development, implementation, and a cognitive evaluation of a definitional question answering system for physicians Journal of Biomedical Informatics, 40(3): pages 236–251, 2007 [85] Xuan Quang Pham, Minh Quang Le and Bao Quoc Ho A Hybrid Approach for Biomedical Event Extraction Proceedings of the BioNLP Shared Task 2013 Workshop, pages 121–124, Sofia, Bulgaria, August 2013 192 Phụ lục A Minh họa số mẫu tài liệu y tế bệnh viện mà luận án sử dụng trình nghiên cứu Các tài liệu trích từ liệu eHealth 2014 SemEval 2015 cung cấp thông qua kho ngữ liệu ShARe, bao gồm tài liệu lâm sàng chọn từ sở liệu lâm sàng MIMIC II (Multiparameter Intelligent Monitoring in Intensive Care II) phiên 2.5 [54] Tài liệu: tóm tắt xuất viện (DISCHARGE_SUMMARY) 16139 |||| 98 |||| 15836 |||| 2015-03-24 00:00:00.0 |||| DISCHARGE_SUMMARY |||| |||| Admission Date: [**2015-03-17**] Discharge Date: [**2015-03-24**] Date of Birth: [**1974-10-03**] |||| |||| Sex: F Service: Neurosurgery HISTORY OF PRESENT ILLNESS: The patient is a 40-year-old female with complaints of headache and dizziness In [**2015-01-14**], the patient had headache with neck stiffness and was unable to walk for 45 minutes The patient also had a similar episode a year and a half ago where she had inability to walk without pain She had a headache at that time which was relieved with Tylenol PAST MEDICAL HISTORY: Hypothyroidism ALLERGIES: Penicillin and Bactrim which causes a rash MEDICATIONS: Levoxyl 1.75 mg 193 PHYSICAL EXAMINATION: On physical examination, her blood pressure was 104/73, pulse 79 In general, she was a woman in no acute distress HEENT: Nonicteric Pupils are equal, round, and reactive to light Extraocular movements are full Pharynx is benign Tongue midline Neck is supple Chest was clear to auscultation Cardiac: S1, S2, regular, rate, and rhythm Abdomen is soft, nontender, nondistended, negative bruits Extremities: No clubbing, cyanosis, or edema Palpable pulses Gait was steady The patient is being admitted status post cerebral aneurysm with clipping The patient underwent right MCA and a right A-comm aneurysm clipping via a right pterional craniotomy without intraoperative complication Postoperative vital signs are stable Patient was afebrile She was awake, alert Pupils are equal, round, and reactive to light EOMs are full No facial droop, no drift Strength down the legs was [**05-18**] On postoperative day #1, the patient was taken to arteriogram, where she underwent a cerebral angiogram to evaluate clipping of the aneurysm It was found at that time that the patient had a third aneurysm and that the MCA aneurysm which was clipped was at the origin of the anterior temporal artery and not the one at the MCA bifurcation The patient was therefore taken back to the operating room for a clipping of the second MCA aneurysm without intraoperative complication The patient was monitored in the Surgical Intensive Care Unit She was alert, awake, oriented, complaining of severe headache, moving all extremities EOMs full, negative drift, a smile was symmetric, IP is [**05-18**] The patient was then discharged to the floor on [**2015-03-20**] She has been out of bed and ambulating with Physical Therapy, tolerating a regular diet, and voiding spontaneously She has been seen by Physical Therapy, and was found to be safe to discharge to home with home physical therapy on [**2015-03-24**] Her incision is clean, dry, and intact and she is neurologically stable 194 DISCHARGE MEDICATIONS: Hydromorphone 2-6 mg po q4h prn Synthroid 175 mcg po q day Colace 100 mg po bid CONDITION ON DISCHARGE: Stable FOLLOW-UP INSTRUCTIONS: She will follow up in 10 days for staple removal with Dr [**Last Name (STitle) 570**] [**Location (un) 571**] [**Initials (NamePattern4) (NamePattern4) **], M.D [**MD Number 572**] **] [**Last Name Dictated By:[**Dictator Info 489**] MEDQUIST36 D: [**2015-03-23**] 08:14 T: [**2015-03-23**] 08:20 JOB#: [**Job Number 1537**] Signed electronically by: DR [**First Name8 (NamePattern2) 571**] [**Name (STitle) **] on: WED [**2015-05-20**] 7:51 AM (End of Report) Báo cáo (ECHO_REPORT) 102920 |||| 176 |||| 2167 |||| 09-23 00:00:00.0 |||| |||| PATIENT/TEST INFORMATION: Indication: Pericardial effusion Height: (in) 68 Weight (lb): 184 195 ECHO_REPORT |||| 2013- BSA (m2): 1.97 m2 BP (mm Hg): 140/80 HR (bpm): 70 Status: Inpatient Date/Time: [**2013-09-23**] at 10:20 Test: TTE (Focused views) Doppler: No doppler Contrast: None Technical Quality: Adequate |||| |||| INTERPRETATION: Findings: LEFT ATRIUM: The left atrium is moderately dilated The left atrium is elongated RIGHT ATRIUM/INTERATRIAL SEPTUM: The right atrium is moderately dilated LEFT VENTRICLE: There is severe symmetric left ventricular hypertrophy The left ventricular cavity size is normal Overall left ventricular systolic function is low normal (LVEF 50-55%) LV WALL MOTION: The following resting regional left ventricular wall motion abnormalities are seen: basal inferoseptal - hypokinetic; mid inferoseptal hypokinetic; basal inferior - hypokinetic; mid inferior - hypokinetic; RIGHT VENTRICLE: The right ventricular free wall is hypertrophied Right ventricular chamber size is normal Right ventricular systolic function is normal AORTIC VALVE: The aortic valve leaflets (3) are mildly thickened MITRAL VALVE: The mitral valve leaflets are mildly thickened There is mild mitral annular calcification TRICUSPID VALVE: Mild tricuspid [1+] regurgitation is seen There is mild pulmonary artery systolic hypertension PERICARDIUM: There is a trivial/physiologic pericardial effusion The effusion appears circumferential Conclusions: 196 Limited focused study The left atrium is moderately dilated The left atrium is elongated 3.The right atrium is moderately dilated 4.There is severe symmetric left ventricular hypertrophy The left ventricular cavity size is normal Overall left ventricular systolic function is low normal (LVEF 5055%) Resting regional wall motion abnormalities include basal and mid septal hypokinesis 5.The right ventricular free wall is hypertrophied Right ventricular chamber size is normal Right ventricular systolic function is normal 6.The aortic valve leaflets (3) are mildly thickened No color doppler study was done to assess for the presence of MR [**First Name (Titles) 4**] [**Last Name (Titles) 15**] The mitral valve leaflets are mildly thickened 8.There is mild pulmonary artery systolic hypertension There is a trivial/physiologic pericardial effusion The effusion appears circumferential An echogenic density in the right ventricle is present consistent with a pacemaker lead Compared with the findings of the prior study (tape reviewed) of [**2013-09-12**], the pericardial effusion is much less The overall function is unchanged Báo cáo X-quang (RADIOLOGY_REPORT) 400001 |||| 587 |||| 22383 |||| RADIOLOGY_REPORT |||| 2012-03-13 13:03:00.0 |||| Q116 CT SINUS/MANDIBLE/MAXILLOFACIAL W/O CONTRAST |||| |||| Clip # 361-6033 Actual report |||| DATE: [**2012-03-13**] 1:03 PM CT SINUS/MANDIBLE/MAXILLOFACIAL W/O CONTRAST [**Clip Number (Radiology) 5189**] Reason: please evaluate for sinusitis Admitting Diagnosis: SPLENOMEGALIA _ _ UNDERLYING MEDICAL CONDITION: 56 year old man with new bilateral ear fullness REASON FOR THIS EXAMINATION: please evaluate for sinusitis No contraindications for IV contrast _ _ FINAL REPORT CLINICAL INFORMATION: Bilateral ear fullness, question sinusitis 197 NONCONTRAST CT OF THE SINUSES AND AXIAL AND CORONAL PROJECTION FINDINGS: The visualized paranasal sinuses are clear with the exception of some slight thickening of central ethmoid cells on the right The osteomeatal units demonstrate soft tissue compromise There appears to be a medial maxillotomy on the right The nasal septum deviates to the left with a rather large spur The cribriform plates are symmetric in height The anterior clinoids are not pneumatized The sphenoid septum inserts on the right carotid groove The mastoid sinuses are clear The nasal pharynx is symmetric in appearance IMPRESSION: Minimal ethmoid sinus thickening Evidence of prior right maxillary surgery Otherwise, negative study Báo cáo điện tâm đồ (ECG_REPORT) 60190 |||| 1982 |||| 19870 |||| ECG_REPORT |||| 2016-09-01 07:43:00.0 |||| |||| |||| |||| Sinus bradycardia Probable prior inferior myocardial infarction Prolonged Q-Tc interval - clinical correlation is suggested for possible in part metabolc/drug effect Since previous tracing of [**2016-08-30**], sinus bradycardia and prolonged Q-Tc interval present 198 ... 146 Hình Kiến trúc tổng quát hệ thống khái thác tài liệu văn lâm sàng .155 Hình Kiến trúc chi tiết hệ thống khai thác tài liệu văn lâm sàng 155 Hình Hệ thống khai thác khái niệm mối quan hệ. .. rút trích mối quan hệ khái niệm tốt  Nghiên cứu đề xuất mơ hình khai thác khái niệm mối quan hệ văn lâm sàng định hướng xây dựng khung phần mềm thống khai thác thông tin văn y tế Hiện có số... hệ thống 1.3.3 Rút trích mối quan hệ Sau rút trích khái niệm bước rút trích mối quan hệ khái niệm Mối quan hệ thực thể y tế quan hệ nhị phân, xét quan hệ cặp đôi hai thực thể Tuy nhiên, mối quan