Naranjo Question Answering using End-to-End Multi-task Learning Model

Naranjo Question Answering using End-to-End Multi-task Learning Model. Rawat, B. P, Li, F., & Yu, H. 25th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD), 2019.
doi abstract bibtex

In the clinical domain, it is important to understand whether an adverse drug reaction (ADR) is caused by a particular medication. Clinical judgement studies help judge the causal relation between a medication and its ADRs. In this study, we present the first attempt to automatically infer the causality between a drug and an ADR from electronic health records (EHRs) by answering the Naranjo questionnaire, the validated clinical question answering set used by domain experts for ADR causality assessment. Using physicians’ annotation as the gold standard, our proposed joint model, which uses multi-task learning to predict the answers of a subset of the Naranjo questionnaire, significantly outperforms the baseline pipeline model with a good margin, achieving a macro-weighted f-score between 0.3652 – 0.5271 and micro-weighted f-score between 0.9523 – 0.9918.

@article{rawat_naranjo_2019,
	title = {Naranjo {Question} {Answering} using {End}-to-{End} {Multi}-task {Learning} {Model}},
	doi = {10.1145/3292500.3330770},
	abstract = {In the clinical domain, it is important to understand whether an adverse drug reaction (ADR) is caused by a particular medication. Clinical judgement studies help judge the causal relation between a medication and its ADRs. In this study, we present the first attempt to automatically infer the causality between a drug and an ADR from electronic health records (EHRs) by answering the Naranjo questionnaire, the validated clinical question answering set used by domain experts for ADR causality assessment. Using physicians’ annotation as the gold standard, our proposed joint model, which uses multi-task learning to predict the answers of a subset of the Naranjo questionnaire, significantly outperforms the baseline pipeline model with a good margin, achieving a macro-weighted f-score between 0.3652 – 0.5271 and micro-weighted f-score between 0.9523 – 0.9918.},
	journal = {25th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD)},
	author = {Rawat, Bhanu P and Li, Fei and Yu, Hong},
	year = {2019},
	pmid = {31799022 NIHMSID: NIHMS1058295 PMCID:PMC6887102},
	pages = {2547--2555},
}

Downloads: 0

{"_id":"bYZ3S3A2hB2ZinWoS","bibbaseid":"rawat-li-yu-naranjoquestionansweringusingendtoendmultitasklearningmodel-2019","author_short":["Rawat, B. P","Li, F.","Yu, H."],"bibdata":{"bibtype":"article","type":"article","title":"Naranjo Question Answering using End-to-End Multi-task Learning Model","doi":"10.1145/3292500.3330770","abstract":"In the clinical domain, it is important to understand whether an adverse drug reaction (ADR) is caused by a particular medication. Clinical judgement studies help judge the causal relation between a medication and its ADRs. In this study, we present the first attempt to automatically infer the causality between a drug and an ADR from electronic health records (EHRs) by answering the Naranjo questionnaire, the validated clinical question answering set used by domain experts for ADR causality assessment. Using physicians’ annotation as the gold standard, our proposed joint model, which uses multi-task learning to predict the answers of a subset of the Naranjo questionnaire, significantly outperforms the baseline pipeline model with a good margin, achieving a macro-weighted f-score between 0.3652 – 0.5271 and micro-weighted f-score between 0.9523 – 0.9918.","journal":"25th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD)","author":[{"propositions":[],"lastnames":["Rawat"],"firstnames":["Bhanu","P"],"suffixes":[]},{"propositions":[],"lastnames":["Li"],"firstnames":["Fei"],"suffixes":[]},{"propositions":[],"lastnames":["Yu"],"firstnames":["Hong"],"suffixes":[]}],"year":"2019","pmid":"31799022 NIHMSID: NIHMS1058295 PMCID:PMC6887102","pages":"2547–2555","bibtex":"@article{rawat_naranjo_2019,\n\ttitle = {Naranjo {Question} {Answering} using {End}-to-{End} {Multi}-task {Learning} {Model}},\n\tdoi = {10.1145/3292500.3330770},\n\tabstract = {In the clinical domain, it is important to understand whether an adverse drug reaction (ADR) is caused by a particular medication. Clinical judgement studies help judge the causal relation between a medication and its ADRs. In this study, we present the first attempt to automatically infer the causality between a drug and an ADR from electronic health records (EHRs) by answering the Naranjo questionnaire, the validated clinical question answering set used by domain experts for ADR causality assessment. Using physicians’ annotation as the gold standard, our proposed joint model, which uses multi-task learning to predict the answers of a subset of the Naranjo questionnaire, significantly outperforms the baseline pipeline model with a good margin, achieving a macro-weighted f-score between 0.3652 – 0.5271 and micro-weighted f-score between 0.9523 – 0.9918.},\n\tjournal = {25th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD)},\n\tauthor = {Rawat, Bhanu P and Li, Fei and Yu, Hong},\n\tyear = {2019},\n\tpmid = {31799022 NIHMSID: NIHMS1058295 PMCID:PMC6887102},\n\tpages = {2547--2555},\n}\n\n","author_short":["Rawat, B. P","Li, F.","Yu, H."],"key":"rawat_naranjo_2019","id":"rawat_naranjo_2019","bibbaseid":"rawat-li-yu-naranjoquestionansweringusingendtoendmultitasklearningmodel-2019","role":"author","urls":{},"metadata":{"authorlinks":{}}},"bibtype":"article","biburl":"http://fenway.cs.uml.edu/papers/pubs-all.bib","dataSources":["TqaA9miSB65nRfS5H"],"keywords":[],"search_terms":["naranjo","question","answering","using","end","end","multi","task","learning","model","rawat","li","yu"],"title":"Naranjo Question Answering using End-to-End Multi-task Learning Model","year":2019}