Research Interests
Health Informatics
My PhD research in health informatics in social network analysis and knowledge transfer/management. I am interested in how SNA can be applied to online communication amongst healthcare practitioners to better understand both how online communication differs from regular forms of communication, and how this information can be leveraged to improve patient care. Working with the
Pediatric Pain Management group at the IWK on a project developing PPM in Thailand, my hope is to leverage online communication forums in order to facilitate tacit knowledge transfer between practitioners.
I am also interested in the inherent knowledege embedded within electonic communications, and how this knowledge can be extracted, organized and leveraged to improve the overall knowledge base. My knowledge linkage project has developed an infobutton to automatically link conversations around clinical problems to published medical literature about that subject.
Other HI subjects I'm interested include incorporating Decision Support Systems into Electronic Medical Records, developing Case Based Reasoning systems that work with EMR data, and on automatic data parsing and data retrieval.
Statistics
As a statistician I am interested in a number of different fields.
computational statistics:
I work almost exclusively with the
R statistical programming language for analysis, and have developed two libraries for it:
longRPart is an extension of the
rpart library for building classification trees for longitudinal outcomes.
ResearchMethods is a set of interactive R programs built using the
tcltk library that help demonstrate the ideas behind the statistical processes we use in a corresponding online course. I have also written a set of help files for creating GUIs in R using the tcltk library, available
here. Note that this is an incomplete list, and not available on CRAN, as I have not contacted the R team about incorporating them into the tcltk library.
literate statistical analysis:
One of the biggest problems with the statistical community is our inability to relate the results of an analysis back to the subject of interest. Using the
Sweave package in R along with LaTeX I attempt to integrate analysis with reporting, with the goal of translating robust statistical analysis directly into useful interpretation.
Publications
- Stewart S. and Sibte Raza Abidi S. (2011). UNDERSTANDING MEDICINE 2.0 - Social Network Analysis and the VECoN System. In Proceedings of the International Conference on Health Informatics, pages 70-79. DOI: 10.5220/0003167100700079 Available Here
- Stewart S, Abidi SS, Finley A. Pediatric pain management knowledge linkages: mapping experiential knowledge to explicit knowledge. Stud Health Technol Inform. 2010; 160: 1184-8. Pubmed Citation
- Davis S, Abidi SS, Stewart S.A compositional personalization approach for designing personalized patient educational interventions for cardiovascular risk management. Stud Health Technol Inform. 2010; 160: 629-33. Pubmed Citation
- Stewart S, Abidi SSR, Finley GA.Linking specialized online medical discussions to online medical literature. In Denecke K et al. (Eds) Using Web Data in the Medical Domain - Proc 1st Intl Workshop on Web Science and Information Exchange in the Medical Web (MedEx 2010). Raleigh, NC.
Available Here
Presentations
- Sam Stewart. Knowledge Linkages: Augmenting Online Clinical Care Discussions with Published Literature. Presented at Med 2.0'10 Conference, Maastricht, NL. November 30, 2010. Slides on slideshare
- Sam Stewart. Social Network Analysis: Understanding Online Communication Patterns. Presented to the Dalhousie Health Informatics Research Network, November 17, 2009.
- Sam Stewart. Linking Electronic Conversations to Pertinent Medical Literature: The Pediatric Pain Mailing List. Presented at the APICS Conference, October 24, 2009.
- Sam Stewart, M. Abdolell. Custom classification trees in R: the longRPart package. Presented at the annual meeting of the Statistical Society of Canada, June 1, 2009, Vancouver BC.
- Also presented at the useR conference , Jully 10, 2009, Rennes, France.
- M. Abdolell, J. Payne, J. Caines, Sam Stewart, W. Lou. An Open-Source Model for Automated Public Health Surveillance Systems: the case of the Nova Scotia Breast Screening Program. Presented at the Dalhousie University Department of Diagnostic Radiology Annual Research Day, April 29, 2008, Halifax, NS.
- Robert Abraham, Mohamed Abdolell and Sam Stewart. Development of an Online IR Teaching Manual with Blackboard Learning System (BLS) and Podcasting. Presented at Society of Interventional Radiology Annual Meeting, March 7-12, 2009, San Diego, CA.
- M. Abdolell, J. Payne, J. Caines, Sam Stewart, W. Lou. An Open-Source Model for Automated Public Health Surveillance Systems: the case of the Nova Scotia Breast Screening Program. Poster presented at the joint annual meeting Statistical Society of Canada & Societe francaise de Statistique, May 27, 2008, Ottawa ON.
- Also presented at the Canadian Public Health Association Annual Conference, June 3 2008, Halifax, NS.
- M. Abdolell, M. LeBlanc, Sam Stewart. Multivariate Regression Trees. Poster Presented at the joint annual meeting Statistical Society of Canada & Societe francaise de Statistique, May 27, 2008, Ottawa ON.
