東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Using Big Data Analytics and Statist...

Davazdahemami, Behrooz.

Linked to FindBook

Google Book

Amazon

博客來

Using Big Data Analytics and Statistical Methods for Improving Drug Safety.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Using Big Data Analytics and Statistical Methods for Improving Drug Safety./
Author:	Davazdahemami, Behrooz.
Published:	Ann Arbor : ProQuest Dissertations & Theses, : 2019,
Description:	127 p.
Notes:	Source: Dissertations Abstracts International, Volume: 81-05, Section: B.
Contained By:	Dissertations Abstracts International81-05B.
Subject:	Bioinformatics. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=13814915
ISBN:	9781088365809

Using Big Data Analytics and Statistical Methods for Improving Drug Safety.
Davazdahemami, Behrooz.

Using Big Data Analytics and Statistical Methods for Improving Drug Safety. - Ann Arbor : ProQuest Dissertations & Theses, 2019 - 127 p.

Source: Dissertations Abstracts International, Volume: 81-05, Section: B.

Thesis (Ph.D.)--Oklahoma State University, 2019.

This item must not be sold to any third party vendors.

This dissertation includes three studies, all focusing on utilizing Big Data and statistical methods for improving one of the most important aspects of health care, namely drug safety. In these studies we develop data analytics methodologies to inspect, clean, and model data with the aim of fulfilling the three main goals of drug safety; detection, understanding, and prediction of adverse drug effects.In the first study, we develop a methodology by combining both analytics and statistical methods with the aim of detecting associations between drugs and adverse events through historical patients' records. Particularly we show applicability of the developed methodology by focusing on investigating potential confounding role of common diabetes drugs on developing acute renal failure in diabetic patients. While traditional methods of signal detection mostly consider one drug and one adverse event at a time for investigation, our proposed methodology takes into account the effect of drug-drug interactions by identifying groups of drugs frequently prescribed together.In the second study, two independent methodologies are developed to investigate the role of prescription sequence factor on the likelihood of developing adverse events. In fact, this study focuses on using data analytics for understanding drug-event associations. Our analyses on the historical medication records of a group of diabetic patients using the proposed approaches revealed that the sequence in which the drugs are prescribed, and administered, significantly do matter in the development of adverse events associated with those drugs.The third study uses a chronological approach to develop a network of approved drugs and their known adverse events. It then utilizes a set of network metrics, both similarity- and centrality-based, to build and train machine learning predictive models and predict the likely adverse events for the newly discovered drugs before their approval and introduction to the market. For this purpose, data of known drug-event associations from a large biomedical publication database (i.e., PubMed) is employed to construct the network. The results indicate significant improvements in terms of accuracy of prediction of drug-evet associations compared with similar approaches.

ISBN: 9781088365809Subjects--Topical Terms:

553671
Bioinformatics.
Subjects--Index Terms:

Adverse Drug Events

Using Big Data Analytics and Statistical Methods for Improving Drug Safety.
LDR:03544nmm a2200409 4500 001 2273621
005 20201109124817.5
008 220629s2019 ||||||||||||||||| ||eng d
020 $a 9781088365809
035 $a (MiAaPQ)AAI13814915
035 $a AAI13814915
040 $a MiAaPQ $c MiAaPQ
100 1 $a Davazdahemami, Behrooz. $3 3551072
245 1 0 $a Using Big Data Analytics and Statistical Methods for Improving Drug Safety.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2019
300 $a 127 p.
500 $a Source: Dissertations Abstracts International, Volume: 81-05, Section: B.
500 $a Advisor: Delen, Dursun.
502 $a Thesis (Ph.D.)--Oklahoma State University, 2019.
506 $a This item must not be sold to any third party vendors.
520 $a This dissertation includes three studies, all focusing on utilizing Big Data and statistical methods for improving one of the most important aspects of health care, namely drug safety. In these studies we develop data analytics methodologies to inspect, clean, and model data with the aim of fulfilling the three main goals of drug safety; detection, understanding, and prediction of adverse drug effects.In the first study, we develop a methodology by combining both analytics and statistical methods with the aim of detecting associations between drugs and adverse events through historical patients' records. Particularly we show applicability of the developed methodology by focusing on investigating potential confounding role of common diabetes drugs on developing acute renal failure in diabetic patients. While traditional methods of signal detection mostly consider one drug and one adverse event at a time for investigation, our proposed methodology takes into account the effect of drug-drug interactions by identifying groups of drugs frequently prescribed together.In the second study, two independent methodologies are developed to investigate the role of prescription sequence factor on the likelihood of developing adverse events. In fact, this study focuses on using data analytics for understanding drug-event associations. Our analyses on the historical medication records of a group of diabetic patients using the proposed approaches revealed that the sequence in which the drugs are prescribed, and administered, significantly do matter in the development of adverse events associated with those drugs.The third study uses a chronological approach to develop a network of approved drugs and their known adverse events. It then utilizes a set of network metrics, both similarity- and centrality-based, to build and train machine learning predictive models and predict the likely adverse events for the newly discovered drugs before their approval and introduction to the market. For this purpose, data of known drug-event associations from a large biomedical publication database (i.e., PubMed) is employed to construct the network. The results indicate significant improvements in terms of accuracy of prediction of drug-evet associations compared with similar approaches.
590 $a School code: 0664.
650 4 $a Bioinformatics. $3 553671
650 4 $a Health sciences. $3 3168359
650 4 $a Artificial intelligence. $3 516317
650 4 $a Public health. $3 534748
650 4 $a Medicine. $3 641104
653 $a Adverse Drug Events
653 $a Adverse Drug Reactions
653 $a Artificial Intelligence
653 $a Data Mining
653 $a Network Analytics
653 $a Pharmacovigilance
690 $a 0715
690 $a 0566
690 $a 0800
690 $a 0564
690 $a 0573
710 2 $a Oklahoma State University. $b Management Information Systems (MS). $3 3551073
773 0 $t Dissertations Abstracts International $g 81-05B.
790 $a 0664
791 $a Ph.D.
792 $a 2019
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=13814915