Extraction of Adverse Drug Effects from Clinical Records E. ARAMAKI* Ph.D., Y. MIURA **, M. TONOIKE ** Ph.D., T. OHKUMA ** Ph.D., H. MASHUICHI ** Ph.D.,K.WAKI.

Slides:



Advertisements
Similar presentations
Applications of one-class classification
Advertisements

Agency for Healthcare Research and Quality (AHRQ)
Pseudo-Relevance Feedback For Multimedia Retrieval By Rong Yan, Alexander G. and Rong Jin Mwangi S. Kariuki
Arnd Christian König Venkatesh Ganti Rares Vernica Microsoft Research Entity Categorization Over Large Document Collections.
Large-Scale Entity-Based Online Social Network Profile Linkage.
Twitter Catches The Flu: Detecting Influenza Epidemics using Twitter Eiji ARAMAKI * Sachiko MASKAWA * Mizuki MORITA ** * The University of Tokyo ** National.
U.S. Food and Drug Administration Notice: Archived Document The content in this document is provided on the FDA’s website for reference purposes only.
Copyright restrictions may apply JAMA Pediatrics Journal Club Slides: Pharmacologic Treatment of Pediatric Headaches El-Chammas K, Keyes J, Thompson N,
Jean-Eudes Ranvier 17/05/2015Planet Data - Madrid Trustworthiness assessment (on web pages) Task 3.3.
Jianwei Lu1 Information Extraction from Event Announcements Student: Jianwei Lu ( ) Supervisor: Robert Dale.
Introduction to Automatic Classification Shih-Wen (George) Ke 7 th Dec 2005.
System Design and Analysis
1 Noun Homograph Disambiguation Using Local Context in Large Text Corpora Marti A. Hearst Presented by: Heng Ji Mar. 29, 2004.
Introduction to Machine Learning Approach Lecture 5.
Stefan Schulz, Thorsten Seddig, Susanne Hanser, Albrecht Zaiß, Philipp Daumke Checking coding completeness by mining discharge summaries.
Results Conclusions Good compliance with writing TTOs however there is room for improvement with adherence to filling in certain information parameters.
Usability Methods: Cognitive Walkthrough & Heuristic Evaluation Dr. Dania Bilal IS 588 Spring 2008 Dr. D. Bilal.
Mining and Summarizing Customer Reviews
Kuang Ru; Jinan Xu; Yujie Zhang; Peihao Wu Beijing Jiaotong University
Title Extraction from Bodies of HTML Documents and its Application to Web Page Retrieval Microsoft Research Asia Yunhua Hu, Guomao Xin, Ruihua Song, Guoping.
Continual Development of a Personalized Decision Support System Dina Demner-Fushman Charlotte Seckman Cheryl Fisher George Thoma.
Empirical Methods in Information Extraction Claire Cardie Appeared in AI Magazine, 18:4, Summarized by Seong-Bae Park.
Adaptive News Access Daniel Billsus Presented by Chirayu Wongchokprasitti.
Health Research & Information Division, ESRI, Dublin, July 2008 The Audit Process.
The use of the Chi-square test when observations are dependent by Austina S S Clark University of Otago, New Zealand.
De-identifying Pathology Reports for Pathology Informatics
Comparative study of various Machine Learning methods For Telugu Part of Speech tagging -By Avinesh.PVS, Sudheer, Karthik IIIT - Hyderabad.
Scott Duvall, Brett South, Stéphane Meystre A Hands-on Introduction to Natural Language Processing in Healthcare Annotation as a Central Task for Development.
Open Health Natural Language Processing Consortium (OHNLP)
Integrated Analyses of Safety Data needed! Marie Louise Valentin, MD Director of Corporate Drug Safety.
Combining terminology resources and statistical methods for entity recognition: an evaluation Angus Roberts, Robert Gaizauskas, Mark Hepple, Yikun Guo.
A Weakly-Supervised Approach to Argumentative Zoning of Scientific Documents Yufan Guo Anna Korhonen Thierry Poibeau 1 Review By: Pranjal Singh Paper.
Math Information Retrieval Zhao Jin. Zhao Jin. Math Information Retrieval Examples: –Looking for formulas –Collect teaching resources –Keeping updated.
PSEUDO-RELEVANCE FEEDBACK FOR MULTIMEDIA RETRIEVAL Seo Seok Jun.
Understanding Medical Articles and Reports Linda Vincent, MPH UCSF Breast SPORE Advocate September 24,
A Scalable Machine Learning Approach for Semi-Structured Named Entity Recognition Utku Irmak(Yahoo! Labs) Reiner Kraft(Yahoo! Inc.) WWW 2010(Information.
Date : 2013/03/18 Author : Jeffrey Pound, Alexander K. Hudek, Ihab F. Ilyas, Grant Weddell Source : CIKM’12 Speaker : Er-Gang Liu Advisor : Prof. Jia-Ling.
Discriminative Dialog Analysis Using a Massive Collection of BBS comments Eiji ARAMAKI (University of Tokyo) Takeshi ABEKAWA (University of Tokyo) Yohei.
Department of Software and Computing Systems Research Group of Language Processing and Information Systems The DLSIUAES Team’s Participation in the TAC.
Combining GATE and UIMA Ian Roberts. University of Sheffield NLP 2 Overview Introduction to UIMA Comparison with GATE Mapping annotations between GATE.
Is avoidable mortality a good measure of the quality of hospital care? Dr Helen Hogan Clinical Senior Lecturer in Public Health London School of Hygiene.
Quality Education for a Healthier Scotland Pharmacy Pharmaceutical Care Planning Vocational Training Scheme: Level = Stage 2 Arlene Shaw Specialist Clinical.
Enhancing Text Classifiers to Identify Disease Aspect Information Rey-Long Liu Dept. of Medical Informatics Tzu Chi University Taiwan.
Number Sense Disambiguation Stuart Moore Supervised by: Anna Korhonen (Computer Lab)‏ Sabine Buchholz (Toshiba CRL)‏
A Classification-based Approach to Question Answering in Discussion Boards Liangjie Hong, Brian D. Davison Lehigh University (SIGIR ’ 09) Speaker: Cho,
How to Write Abstract How to write title? a good title (typically 10–12 words long) 6,7 will use descriptive terms and phrases that.
Support Vector Machine Based Orthographic Disambiguation Eiji ARAMAKI, Takeshi IMAI, Kengo MIYO, Kazuhiko OHE Hospital “center” and “centre” are equivalent?
Detection of Spelling Errors in Swedish Clinical Text Nizamuddin Uddin and Hercules Dalianis Department of Computer and Systems Sciences, (DSV)
Automatic Discovery and Processing of EEG Cohorts from Clinical Records Mission: Enable comparative research by automatically uncovering clinical knowledge.
Is avoidable mortality a good measure of the quality of healthcare? Dr Helen Hogan Clinical Senior Lecturer in Public Health London School of Hygiene and.
Virtual Examples for Text Classification with Support Vector Machines Manabu Sassano Proceedings of the 2003 Conference on Emprical Methods in Natural.
Safety in Practice Learning Session 3 PHO and Facilitator: WPHO – Andrew Jones Team members: Kirsty Laws, Allie Waretini, Mel Lanz, James Recordon Silverdale.
Statistical Criteria for Establishing Safety and Efficacy of Allergenic Products Tammy Massie, PhD Mathematical Statistician Team Leader Bacterial, Parasitic.
Consumer Health Question Answering Systems Rohit Chandra Sourabh Singh
Methods We employ the UMLS Metathesaurus to annotate ICD-9 codes to MedDRA preferred terms (PTs) using the three-step process below. The mapping was applied.
Multi-Class Sentiment Analysis with Clustering and Score Representation Yan Zhu.
Opinion spam and Analysis 소프트웨어공학 연구실 G 최효린 1 / 35.
Best-of-Breed Hybrid Methods for Text De-identification Yang H, Garibaldi JM. Automatic detection of protected health information from clinical narratives.
Faculty of Information Technology, Brno University of Technology, CZ
Constructing a Predictor to Identify Drug and Adverse Event Pairs
Showcasing work by Jonnageddala, Liaw, Ray, Kumar, Chang, and Dai on
Medication Information Extraction
Online Conditional Outlier Detection in Nonstationary Time Series
CRF &SVM in Medication Extraction
مدیریت داده ها و اطلاعات آزمایشگاه پزشکی
Eiji Aramaki* Sadao Kurohashi* * University of Tokyo
Text Categorization Document classification categorizes documents into one or more classes which is useful in Information Retrieval (IR). IR is the task.
Sampling Sampling is choosing a sample.
Sampling Sampling is choosing a sample.
Model Enhanced Classification of Serious Adverse Events
Presentation transcript:

Extraction of Adverse Drug Effects from Clinical Records E. ARAMAKI* Ph.D., Y. MIURA **, M. TONOIKE ** Ph.D., T. OHKUMA ** Ph.D., H. MASHUICHI ** Ph.D.,K.WAKI * Ph.D. M.D., K.OHE * Ph.D. M.D., * University of Tokyo, Japan ** Fuji Xerox, Japan Our material is Discharge Summary

Background The use of Electronic Health Records (EHR) in hospitals is increasing rapidly everywhere They contain much clinical information about a patient’s health BUT  Many Natural Language texts ! BUT  Many Natural Language texts ! Extracting clinical information from the reports is difficult because they are written in natural language

NLP based Adverse Effect Detecting System We are developing a NLP system that extracts medical information, especially Adverse Effect, form natural language parts INPUT – a medical text (discharge summary) OUTPUT – Date Time – Medication Event – Adverse Effect Event ≒ i2b2 Medication Challenge But our target focuses only on adverse effect Adverse Effect Relation (AER)

Why Adverse Effect Relations? Clinical trials usually target only a single drug. BUT: real patients sometimes take multiple medications, leading to a gap separating the clinical trials and the actual use of drugs For ensuring patient safety, it is extremely important to capturing a new/unknown AEs in the early stage.

DEMO is available on

副作用関係の推定 System Demo

CcCc 副作用関係の推定 System Demo has no complications at the time of diagnosis 6/23-25 FOLFOX6 2 nd. 6/24, 25: moderate fever (38 ℃ ) again. a fever reducer…. Adverse Effect Medication Relation

The point of This Study (1) Preliminary Investigation: How much information actually exist? – We annotated adverse effect information in discharge summaries (2) NLP Challenge: Could the current NLP retrieve them? – We investigated the accuracy of with which the current technique could extract adverse effect information

Outline Introduction Preliminary Investigation – How much information actually exist in discharge summary? NLP Challenge Conclusions

Material & Method Material: 3,012 Japanese Discharge Summaries 3 humans annotated possible adverse effects due to the following 2 steps Lasix for hypertension is stopped due to his headache. Step 1 Event Annotation Step 2 Relation Annotation XML tag = Event XML attribute = Relation

Annotation Policy & Process We regard only MedDRA/J terms as the events. We regarded even a suspicion of an adverse effect as positive data. Entire data annotation is time-consuming → We split data into 2 sets SET-A (Event Rich parts): contains keywords such as Stop, Change, Adverse effect, Side effect SET-B: The other adverse effect terminology Full annotated Randomly sampled & annotated

14.5%×53.5% %×11.3% = 17.4% SET-B SET-A

Results of Preliminary Investigation About 17% discharge summaries contain adverse effect information. – Even considering that the result includes just a suspicion of effects, the summaries are a valuable resource on AE information. We can say that discharge summaries are suitable resources for our purpose.

Outline Introduction Preliminary Investigation NLP Challenge – Could the current NLP technique retrieve the AEs? Conclusions

Combination of 2 NLP Steps 2 NLP steps directly correspond to each annotation step Lasix for hyperpiesia is stopped due to the pain in the head. symptom Medication Adverse Effect Relation Event Annotation Relation Annotation ≒ Named Entity Recognition Task = Relation Extraction Task, which is one of the most hot NLP research topics.

Step1: Event Identification Machine Learning Method – CRF (Conditional Random Field) based Named Entity Recognition Feature – Lexicon (Stemming), POS, Dictionary based feature (MedDRA), window size=5 Material – SET-A Corpus with Event Annotations state-of-the-art method at i2b2 de-identification task Standard Feature Set

Step1: Result of Event Identification Result Summary Cat. of EventPrecisionRecallF-measure Medication Event AE Event All accuracies (P, R) >> 80 %, F>0.80, demonstrating the feasibility of our approach Considering that the corpus size is small (435 summaries), we can say that the event detection is an easy task

Step2: Relation Extraction Method Basic Approach ≒ Protein-Protein Interaction (PPI) task [BioNLP2009-shared Task] Example Lasix for hypertension is stopped due to his headache For each m (Medications) For each a (Adverse Effects) judge_it_has_rel (a, m) For each m (Medications) For each a (Adverse Effects) judge_it_has_rel (a, m) (1) judge_it_has_AER (Lasix, hypetension) (2) judge_it_has_AER (Lasix, headach)

(1) PTN-BASED: heuristic rules using a set-of- keyword & word distance..is on ACTOS but stopped for relief of the edema. n=1 keyword n=4 Judge_it_has_AER (m, a, keyword=stopped, windowsize5) (2) SVM-BASED: Machine learning approach – Feature: distance & words between two events ( medication & adverse effect) Two judgment methods See proceedings for detailed

Step2: Result of Relation Extraction PrecisionRecallF-measure PTN-BASED41.1%91.7% %62.3%0.598 SVM-BASED Both PTN & SVM accuracies are low (F<0.65) → the Relation extraction task is difficult! SVM accuracy is significant (p=0.05) lower than PTN (1) Corpus size is small (2) positive data << negative data Machine learning suffers from such small imbalanced data

Outline Introduction Preliminary Investigation NLP Challenge Discussions – (1) Overall Accuracy – (2) Controllable Performance – (3) Event Distribution Conclusions

Discussion (1/3) Overall Accuracy The overall accuracy is estimated by the combined accuracies of step1 & step2 Overall (= step1 × step2) Precision0.289 (=0.855 × × 0.390) Each NLP step is not perfect, so, the combination of such imperfect results leads to the low accuracy (especially many false positives; low precision) Recall (=0.802 × × 0.917)

Discussion (2/3) Performance is Controllable Precision & Recall curve in SVM The performance balance between recall & precision could be controlled High precision setting High recall setting That is a strong advantage of NLP

Discussion (3/3) Event Distribution We investigated the entire AE frequency for each medication category. distribution acquired from annotated real data distribution acquired from our system results AE freq. distribution of Drug #1

Discussion (3/3) AER Distribution Then, we checked the goodness of the fit test, which measures the similarity between two distributions Med. 1 Med. 2 Med. 3 Med. 4 Med. 5 Total P-value High p-value (p=0.011 > 0.01) indicates two distributions are similar.

Outline Introduction Preliminary Investigation NLP Challenge Discussions Conclusions

Conclusions (1/2) Preliminary Investigation: – About 17% discharge summaries contain adverse effect information. – We can say that discharge summary are suitable resources for AERs NLP Challenge: – Could NLP retrieve the AE information? – Difficult! Overall accuracy is low

Conclusions (2/2) BUT: 2 positive findings: (1) We can control the performance balance (2) Even the accuracy is low, the aggregation of the results is similar to the real distribution IN THE FUTURE: – A practical system using the above advantages – More acute method for relation extraction

Thank you Contact Info – Eiji ARAMAKI Ph.D. – University of Tokyo – –