JAIST Repository >
School of Knowledge Science >
Articles >
Journal Articles >

Please use this identifier to cite or link to this item: http://hdl.handle.net/10119/15476

Title: Sequence-based Measure for Assessing Drug-Side Effect Causal Relation from Electronic Medical Records
Authors: Dang, Tran Thai
Ho, Bao Tu
Keywords: Drug-Side Effect
Causal Relation
Electronic Medical Rec
Issue Date: 2017-10-17
Publisher: Springer
Magazine name: Communications in Computer and Information Science
Volume: 780
Start page: 53
End page: 65
DOI: 10.1007/978-981-10-6989-5_5
Abstract: The recent prevalence of electronic medical records offers a new way to detect the drug-side effect causality. However, this approach faces with the problem of identifying the likely causal relation between drugs and side effects in a huge space of possible relations in which many relations are not causal ones, but frequently observed. In existing work, the likely causal relation is almost detected by using frequency-based measures of drug-side effect pair co-occurrence, but the accuracy is rather low due to the frequent co-occurrence of non-causal pairs. Our key as- sumption on the causality of the drugs and side effects is that the causal- ity occurring at a medication event has an association with the therapy history of this event. The assumption is employed as a constraint in the proposed sequential-based model named Medication Therapy Progress- based Model (MTPM). Experiments show a significant improvement of accuracy from 4% to 9% when comparing MTMP and existing methods, as well as reflect the likelihood of the assumption.
Rights: This is the author-created version of Springer, Tran-Thai Dang, Tu-Bao Ho, Communications in Computer and Information Science, 780, 2017, 53-65. The original publication is available at www.springerlink.com, http://dx.doi.org/10.1007/978-981-10-6989-5_5
URI: http://hdl.handle.net/10119/15476
Material Type: author
Appears in Collections:a10-1. 雑誌掲載論文 (Journal Articles)

Files in This Item:

File Description SizeFormat
23963.pdf390KbAdobe PDFView/Open

All items in DSpace are protected by copyright, with all rights reserved.

 


Contact : Library Information Section, Japan Advanced Institute of Science and Technology