1. Analytics Studies in Health Care
Introduction
Brian Fisher
bfisher@sfu.ca
SFU Interactive Arts & Technology / Cognitive Science
UBC Media & Graphics Interdisciplinary Centre
2. My Background
• BA Biology, Medical Biophysics at CWRU Med
• Scientific Programmer, Varian
• Ph.D Experimental Psychology, UCSC
3. Currently
• SFU School of Interactive Arts and Technology
• Social Cognition and Interactive Expertise in Natural and
Computational Environments (SCIENCE lab)
• UBC Media & Graphics Interdisciplinary Centre
(MAGIC)
• Enable non-tech units to do (and fund) tech projects
• New ventures: Thoughtshare, Xing Xing
• Not-for profits: New Ventures BC
• Education: HCI grad cert, VA training
4. SCIENCElab Applications
• Emergency Management
• Mobile analytics / sensor analytics
• “Virtual EOC” visual analytic environment
• Aircraft Safety, Reliability
• “Pair analytics” of complex quant and text data
• Healthcare Monitoring & Management
• Complex data in health research (CFRI)
• Public health monitoring & management (BC Injury
Research and Prevention Unit)
• Government policy
• World Bank projects with Vicki Lemieux
5. Recent SCIENCElab people
• Dr. Richard Arias-Hernández
• Dr. Linda Kaastra
• Dr. Samar Al-Hajj
• Nadya Calderón
• Tera Marie Green
• Ethan Soutar-Rau
6. 1995 InfoVis Conference Keynote
"Information Visualization: Wings for the Mind"
• Increasing the memory and processing
resources available to users
• Reducing the search for information by using
visual representations to enhance the
detection of patterns
• Engaging perceptual inference operations
• Using perceptual attention mechanisms for
monitoring
• Supporting manipulation of information
(Card, 1995)
7. “To me that looks like a plate of
spaghe1… Your users are bears”
A re9red police officer
Microso> Research/Na9onal Visualiza9on
and Analy9cs Center Workshop on Regional
Preparedness, 2007
7
9. "Il n'existe pas une catégorie de sciences auxquelles on puisse donner le nom
de sciences appliquées. Il y a la science et les applications de la science, liées
entre elles comme le fruit à l'arbre qui l'a porté” ——— Louis Pasteur
Pure Basic
Research
(Bohr)
Use-inspired
Basic
Research
(Pasteur)
Sampling,
Description,
Taxonomy
(Audubon)
Pure Applied
Research
(Edison)
Quest for
Fundamental
Understanding?
No
Yes
Consideration of Use ?
No
Yes
(1822–95)
Pasteur’s Quadrant (Stokes)
10. How can science help vis?
• Design system in accordance with known
visual cognition theories & phenomena
• Evaluate prototypes & systems with regard to
theories of visual cognition:
• Iterate
• “Close reading” of interface &/or video of interface use
• Look for threats/abilities human Cognitive Architecture
• Formulate hypothesis, do lab experiment on CA
• Change design, evaluate in field study
11. Example: Visual Cognition in ATC
• “Close reading”
NextGen ATC
fishtank video
• Identify threat to
human cognitive
architecture
• Here: How will global
motion cues affect
FINSTs?
Liu, G. Austen, E. L., Booth, K.S. Fisher, B., Argue, R. Rempel, M.I., & Enns, J.
(2005) Multiple Object Tracking Is Based On Scene, Not Retinal, Coordinates.
Journal of Experimental Psychology: Human Perception and Performance. 31(2),
Apr 2005, 235-247.
http://www.youtube.com/watch?v=tKJVB4id_TY
14. Origins of Visual Analytics
Tools support understanding implications of data
▪ Synthesize information & derive insight from massive, dynamic,
ambiguous, & conflicting data
▪ Detect the expected & discover the unexpected
▪ Build timely, defensible, & understandable assessments
▪ Communicate assessments effectively for action.
“The beginning of knowledge is the discovery of
something we do not understand.”
~Frank Herbert (1920 - 1986)
Jim Thomas slide
“The science of analytical reasoning facilitated
by interactive visual interfaces”
15. “A grand challenge for the scientific
enterprise”
• Battelle/PNNL 2004 R&D Agenda panel
• University: Brown, GMU, Georgia Tech, OSU, Penn State,
Purdue, SFU , Stanford, UC, UI, UM, UNC, UU, WPI
• Industry: Boeing, Microsoft, PARC, Sandia Labs
• Gov: CIA, DHS, FBI, NIST, NSA, unspecified
• National Visualization and Analytics Center
• VA Industry Consortium
• DHS University Center of Excellence in VA
• IEEE VAST conference
• NSF FODAVA
• EU 7th Framework VisMaster
Coordination Action
17. VA as a Translational
Cognitive Science
• Lab: Directed basic research in cognition
• Clinic: Analysis “in the wild”— what validates it?
• Lab to Clinic translation
• Design & software engineering methods that incorporate
scientific knowledge
• Field experiments to evaluate theory in context
• “Clinic” to “Lab” translation
• Cognitive ethnography (Hutchins) ex: Emergency ops
• “Close reading” of interface using cogsci theory
• Pair Analysis field experiments w JAT video analysis
18. Research Methods
• Cognitive ethnography
• Document current tools and methods
• Understand “lab culture” e.g goals & reward structures
• “Pair analytics” with VA tools & datasets,
“close reading” visualization & collaborative
“joint activity” of analysis
• Collaborate with UNC Charlotte Vis Center on
design and engineering of interactive Principle
Component analysis (iPCA) tool
19. Ongoing Directed Basic
Research
• Perceptual cognition (e.g. FINSTs & ATC)
• Human Cognitive Models with Bill Ribarsky
and Daniel Keim
• Personal Equation work (Po, Greensmith) to
capture innate and acquired differences
• Eye movements & visual cognition in complex
visualization environments with Daniel
Weiskopf and colleagues at Stuttgart
20. • “Pair Analysis” sessions
designed for JAT video analysis
• Student visual analysis (VAE)
& subject matter expert (SME)
• VAE drives, expert navigates
Visual Analytics in Pair Analysis
Arias-Hernandez, R, Kaastra, L.T., and Fisher, B. (2011) Joint Action Theory and Pair Analytics: In-vivo Studies of
Cognition and Social Interaction in Collaborative Visual Analytics. In L. Carlson, C. Hoelscher, and T. Shipley
(Eds.), Proceedings of the 33rd Annual Conference of the Cognitive Science Society (pp. 3244-3249). Austin TX:
Cognitive Science Society.
•Video analysts & capture screen
•Video analysis focuses on understanding of
process analytics
•Video analysis tools: Chronovis*, NVIVO, Atlas.ti
21. Joint Activity Theory (Clark)
• Communication as Joint Activity, like
playing a duet or paddling a canoe
• H. H. Clark’s theory of joint activity
• Defines kinds of common ground
• Formalizes the notion of activity as a “joint
action”
• Describes the processes by which common
ground is developed through joint action
• We extend Clark from F2F to focus on
technology as integral to communication
and collaborative analysis
23. Reduce Child Mortality from
Infectious Diseases
• UBC & BC Children’s Hospital (Dr. Kollmann),
Stellenbosch Inst. for Advanced Studies
• Morbidity/mortality in rural South African children
• Effects of public health intervention e.g vaccination
• We build on UBC BioSci/CompSci student
David Shih, Cogsys student Kevin Ho on work
terms (Rensink)
• Goal is to understand factors that enable
competent immune response to infection
24. Data: Flow cytometry (FlowJO)
• Compare 2 blood samples (e.g. before/after vaccine)
• Look at 18 cytokines that reflect WBC activation x 3M
White Blood Cells
• Percent positive for given cytokine
• MFI (Mean Fluorescence Intensity) for activation
• 3 cell types
• Monocytes
• mDC
• pDC
25. Construct: PolyFunctionality Degree
(PFD)
• A measure of the degree to which a cell is
polyfunctional, i.e. production of multiple cytokines
• Expressed in terms
of percentage
• Pfd1 = % 1 cytokine
• Pfd2 = % 2 cytokines
• Pfd3 = % 3+ cytokines
27. • Does the increase in
polyfunctionality in
the adult simply
reflect increased
overall cytokine
production, or is it a
unique functional
response the
neonate is less
capable of? PFD1 PFD2
PFD3+
Research Question
30. Conclusions & future work
• Best Student Discovery Exhibition Award
(IEEE VAST 2011)
• iPCA approach shows promise
• iPCA needs cognitive and perceptual skill
training to enable health professionals to
synthesis meaningful information.
• Future collaboration depends on
• Buy-in of health science researchers and technology
developers
• Appreciation of the need for science-informed design
and evaluation
31. Public Health Example:
Child Injury Prevention
Samar Al-Hajj
SIAT Ph.D. student
Al-Hajj, S., Pike, I., Riecke, B., & Fisher, B. (2013) Visual Analytics for Public Health: Supporting Knowledge
Construction and Decision-Making (full paper). Proceedings of the 46th Annual Hawaii International Conference
on System Sciences. IEEE Digital Library
Collaborative Visual Analytics for Public Health: Facilitating Problem Solving and Supporting Decision-Making
(2014 SFU Doctoral dissertation)
Dr. Samar Al-Hajj
Dr. Ian Pike
UBC Pediatrics
32. BC Child Injury Project
• Apply VA methods to multi-stakeholder
exploratory analysis of injury indicator data.
• Data are complex and heterogeneous (patients’ age,
sex, socioeconomic status as well as injuries’ types
and geographic locations).
• Visualizations created using Tableau Software.
• Build & evaluate collaborative VA to facilitate
insight generation, knowledge construction
and decision-making.
36. Methods
• 6 Analytical Sessions: Visual Analysis and
Subject Matter Experts Pair Analysis
• 1 Group Analysis Session: VAE, multiple SME,
Process facilitator
• JAT analysis
• Questionnaire document perceived utility &
usability of tech and process
• Goal is to develop and evaluate a socio-
technical system for making group decisions
that are informed by data
39. Challenges
• Software engineering challenges
• Funding challenges – NSERC or CIHR?
• Understanding cultures
• The laboratory
• The larger research community
• Interdisciplinary research management
• Responsibility for the project?
• Managing expectations
• Insuring research products for all
40. What we need to succeed
Success in big-tent visual analytics will depend
on ongoing collaboration on a set of related
projects in application-focused institutes,
supported by an interdisciplinary community of
visual analytics researchers.
42. The PMI is incorporated as a not-for-profit Canadian company based in
Vancouver, BC
Board of Directors: Pieter Cullis (Co-Chair), David Huntsman (Co-Chair), Bruce
McManus, Jim Russell and Martin Dawes
COO: Rob Fraser
PMI Vision: Introduce an individualized approach to preventive and curative
healthcare based on the molecular makeup of the individual and their disease.
43. The PMI Group of Companies
PBI
1) GenXys: PGx to guide prescription practices in family practice
▪ Principals: M. Dawes
▪ Status: Operating
2) Contextual Genomics: Cancer gene analysis to guide
chemotherapy
▪ Principals: D. Huntsman, C. Wagner
▪ Status: Operating
3) Cyon Therapeutics: Treatments for septic shock
▪ Principals: J. Russell
▪ Status: Operating
4) Personalized Biomarkers Inc: Omics to guide diabetes therapy
▪ Principals: T. Elliot
▪ Status: Seed
5) Microbiome Insights: Microbiome analyses
▪ Principals: B. Finlay, B. Mohn, M. Kendall
▪ Status: Seed, but operational
6) Molecular You Corporation: Longitudinal Omics Analysis
▪ Principals: P. Cullis, R. Fraser
▪ Seed, not operational
44. Engaging the PMI
• Work on existing projects, collaborate w
UBC, ECUAD, others: what is our best role?
• Focus on scalability, system integration
• Engage PMI in personalized wellness and
health promotion, opportunities for tech
development
• Consider environmental determinants of
health and wellness with Canadian
Environmental Health Atlas Bruce Lanphear