Dr. Uli Muellner - Lab SAVI

50 %
50 %
Information about Dr. Uli Muellner - Lab SAVI

Published on October 31, 2018

Author: trufflemedia

Source: slideshare.net

1. LAB/savi – supercharging VDL data for better animal health decision-making Allen D. Leman Swine Conference 17 Sept 2018

2. 2 Acknowledgments VDL team • Stephanie Rossow • Jerry Torrison • Mary Thurn • Paulo Fioravante • Albert Rovira • Matt Sturos • Veterinarians participating in stakeholder workshop June 2018

3. Working in a challenging data environment • Lab data is complex, hierarchical, highly diverse… • Data reaching back to early 2001 • Over 500 different procedures • Granular access to cases, linked to a network of affiliates • Producer • Owner • Submitter • Pathologists • Vets, vet clinics • Pharmaceutical clients 3

4. 4 VDL goals • Protecting animal and human health through early detection and monitoring of animal disease • Provide “information for action” for its stakeholders • Easy access for customers

5. Planned solution: LAB/savi Currently available: • Online access to case reports and results New features in LAB/savi: • Combining results to generate insights into population-level trends • Better ways to interrogate data (i.e. spatial exploration) • Add-ons: genetic analysis, susceptibility profiles to investigate antimicrobial resistance 5

6. Interface is key • Needs to be easy to use! • Serving different audiences • Providing information you need at the right time! • “No more than three clicks away…” 6

7. Some inspiration… 7

8. LAB/savi in action...

9. Demo: Filters demo data only

10. Demo: Case selection on map demo data only

11. Demo: Map interactions demo data only

12. Genetics

13. Mock-up – demo data only

14. Mock-up – demo data only

15. Mock-up – demo data only

16. Mock-up – demo data only

17. Susceptibility Profiles

18. Mock-up – demo data only

19. Mock-up – demo data only

20. Mock-up – demo data only

21. Mock-up – demo data only

22. Technology

23. Technology • Extensive use of statistical programming language R • Use of R packages for data visualisation and specialist analysis • R Shiny for publishing to the web o Extended by JavaScript and web technologies where required • “Reactive” programming • Rapid development • Ability to scale > dedicated server at UMN 23

24. Shiny Server Connect PostGreSQL Database 24 API Cases Procedures Results Users & access VDL clients Veterinarians Researchers Laboratory Information System (LIMS) Veterinary Diagnostic Lab (VDL) Infrastructure

25. 25 Access from anywhere

26. Outlook • Prototype testing with early adopter (Nov/Dec 18) • Fully functional first version – mid 2019 Going forward: • High scalability • Can extend to include all procedures completed at the VDL • Target additional species: e.g. poultry • Analytical scalability: risk maps; advanced quantitative epidemiological models • Further work with stakeholders to identify additional information needs that LAB/savi can support 26

27. THANK YOU! VDL project team Stephanie Rossow Jerry Torrison Mary Thurn Paulo Fioravante Albert Rovira Matt Sturos EPI-interactive team Petra Muellner Christina Ahlstrom Liang Yang Shanna Tervoort-McLeod Anna Poulin uli@epi-interactive.com 27

Add a comment