The Biomedical Informatics Program (BIP) aims to maximize the scientific impact of research proposals through four specific aims: 1) Developing interoperable applications and repositories to integrate multi-scale data across institutions. 2) Consulting with investigators to coordinate use of informatics tools. 3) Educating researchers on biomedical informatics principles and tools. 4) Developing novel informatics techniques for large data integration, semantic data extraction/transformation/loading, and testing data integrity in federated environments. The BIP brings together researchers from Emory, Morehouse, and Georgia Tech to provide resources and training to enable collaborative, data-driven translational research across ACTSI institutions.

1. Biomedical Informatics Program

2. BIP Team
 Emory  Morehouse  Georgia Tech
 Joel Saltz  Elizabeth Ofili  Barbara Boyan
 Tahsin Kurc  Alex Quarshie  Mary Jean Harrold
 Tim Morris  Adam Davis  Alessandro Orso
 Marc Overcash  Doug Blough
 Andrew Post  Karsten Schwan
 Sanjay Agravat  Mustaque Ahamad
 Circe Tsui
 Tony Pan
 Ashish Sharma
 Fusheng Wang
 Carlos Moreno

3. BIP Objectives
The overarching goal of the ACTSI Biomedical Informatics Program (BIP) is to
maximize the scientific impact of ACTSI investigator proposals and facilitate
novel translational research by 1) enabling management, linkage, analysis and
mining of multi-scale, multi-dimensional data across ACTSI institutions and 2)
training, consulting, and assisting ACTSI investigators for more effective
application of bioinformatics, biostatistics, and informatics in their projects.

4. BIP Aims
 Specific Aim 1: Develop a suite of interoperable, linked applications and
repositories for management and integration of clinical, "omics", imaging,
laboratory, and tissue data.
 Specific Aim 2: Engage ACTSI investigators via consultations to maximize the
impact of ACTSI investigator proposals through coordinated use of bioinformatics,
biostatistics, and informatics.
 Specific Aim 3: Educate researchers and others in our academic community on the
principles and best practices of biomedical informatics and use of biomedical
informatics applications and tools. Closely coordinated with the Research
Education, Training, and Career Development program.
 Specific Aim 4: Develop novel biomedical informatics techniques and tools for: 1)
synthesis of information from very large multi-scale, multi-dimensional data, 2)
tools to create patient data registries through a semantic extract-transform-load
process, and 3) methods and tools for testing data integrity and maintaining
security in federated environments.

5. Aim 1: Develop suite of interoperable, linked applications
 Develop integrative, federated ACTSI virtual
information warehouse
 Integrated clinical/imaging/”omic”/biomarker/tissue
information should always be available
 A virtually centralized, Atlanta wide information
warehouse that has all relevant data
 Index and federate information generated throughout
ACTSI -- information available from patients seen and
information gathered at any ACTSI site, specimens sent to
any affiliated core, imaging carried out at any affiliated site
 Governance and technology to manage authentication,
authorization

6. Applications and Databases Deployed by BIP
Application Deployed Content and Usage Function and Benefits
CR-Assist Dec 2005 322 studies since deployment, 126 active Enables researchers to manage
studies, 4258 participants, and 439 users participants, schedule and track study
events (visits, laboratory tests), and print
labels for bio-specimen collection.
eBIRT Jul 2010 424 services from 57 cores Enables discovery of relevant laboratories,
expertise, and services.
PAIS Database Aug 2010 In silico brain tumor study database of Enables researches to store, index, and
image analysis results from 307 slides explore image markups and annotations on
micro-anatomic structures for correlative
studies.
REDCap Apr 2010 1057 data instruments created, used by 75 Provides support for researchers to easily
studies capture and manage clinical research data.
Nautilus LIMS Jul 2010 421 studies in LIMS and 94 users; number Facilitates structured and more efficient
of aliquots received in LIMS: 754,099 management of laboratory workflows and
bio-specimens via a common
infrastructure.
AIW Clinical Registry Mar 2011 5 years of data on 4061 Emory patients Provides semantically annotated, easy-to-
query databases of clinical data for clinical
research.
AIW-Readmissions May 2011 5 years of clinical and administrative data on Provides a semantically annotated dataset
149,814 Emory patients for analyzing hospital readmissions.
MSM i2b2 Oct 2010 Clinical information from 21,000 patients Easy to use interfaces for researchers to
access EHR data for cohort identification.

7. Example Translational Research Projects
 In silico study of brain tumors
 Integrative analysis of image, omics, and clinical outcome data
 Cardiovascular Studies
 Correlative analyses of integrated data from databases of clinical
information as well as genomic and phenotypic information
 Minority-Health Grid (MH-GRID)
 Advance genomic science and personalized medicine in minority
health research
 Big Bethel AME Project
 Uses principles of community engaged research and biomedical
informatics tools to assist diabetic congregants of the Big Bethel AME
church in Atlanta to improve diabetes self management skills.

8. Example Translational Research Projects
 Glenn Project
 Increasing rate of consent and research specimen collection at Emory
University Hospital, Emory Midtown Hospital and Grady Hospital
 Early hospital readmission
 Understand relationship between disease conditions, treatments and
environmental factors in predicting hospital readmissions within 30
days.
 Clinical Interaction Network
 Search and analysis of de-identified patient data to help investigators
plan studies
 CIN obtains real time notification when study patient is hospitalized
and obtains real time EMR data

9. In Silico Brain Tumor Research Center
(ISBRTC)
 A research center of excellence for in silico study of brain tumors
 Systematically execute in silico analyses (experiments) using
complementary data types
 Collaborative effort among four institutions
 Emory University
 Thomas Jefferson University
 Henry Ford Hospital
 Stanford University
 Initial focus on gliomas
 Better Classification
 Study Biology of Progression
 Development of Methods and Workflows
 “Companion” National Library of Medicine R01 funded, additional
companion proposals in review and preparation
7/9/2012
9

10. Minority-Health Grid (MH-GRID)
 PI: Gary Gibbons, multi-site project involving
Morehouse School of Medicine, Grady, Jackson Hinds
Clinic, and Kaiser
 Health disparities research focusing on hypertension
in minority populations
 Integration of de-identified clinical phenotypes,
social-environmental data elements, biospecimens,
laboratory data, and genomic information
 Data sharing and federation infrastructure will build
on the BIP Architecture and the Enhanced Registries
system

11. Big Bethel AME Project
 PI: Priscilla Igho-Pemu. A Pilot study involving CIN, BIP, and CER programs of the
ACTSI and Big Bethel AME Church.
 Hypothesis: Diabetic patients who use ehealthystrides and its social networking
forum will demonstrate better diabetes self management skills.
 Main outcome variable: attainment of at least 3/7 of the American Association of
Diabetes Educators (AADE7) diabetes self-care behavioral goals.
 Uses principles of community engaged research and biomedical informatics tools
to assist consented diabetic congregants (Participants) of the Big Bethel AME
church in Atlanta under the guidance of a trained coach to improve diabetes self
management skills.
 Supports consumers as drivers of health transformation.
 Coaches and participants are trained on the use of the ehealthystrides
application, personal health record creation, AADE7 goals and use of the
structured behavioral goal setting tools.
 A community access “kiosk” with internet access and web portal has been
provided at the Big Bethel AME church premises to enhance training and
utilization of informatics tools by participants.
 110 participants have currently been enrolled.

12. GLENN Project
 POC: Dan Brat, project to define streamlined processes and
systems for Breast Cancer bio-banking at Winship
 Primary goal: Increasing rate of consent and research
specimen collection at Emory University Hospital, Emory
Midtown Hospital and Grady Hospital
 Integration of identified and de-identified clinical phenotypes
with available specimens for use in research
 Architecture will utilize ACTSI master study participant index,
enterprise LIMS implementation and Emory enterprise service
bus
 Generic architecture for use to support bio-banking across
Emory/ACTSI

13. LIMS
 Establish a ‘virtual’ biobank and specimen tracking
infrastructure across the ACTSI.
 Labs at many of our Clinical Interaction Networks are
in deployment or close to deployment:
 Emory University, Morehouse School of Medicine, Grady,
Midtown, and Children’s
 In process for next phase laboratories:
 Hope Clinic
 Neurology
 Children’s Research Laboratories

14. Topic-specific Clinical Registries
 Created using AIW infrastructure
 Novel semantic extract-transform-load (ETL) tool in AIW to
create disease specific, semantically annotated clinical
repositories
 i2b2 is used as user-facing presentation layer
 Multiple registries are in various stages of
development for cardiovascular disease, diabetes,
oncology, and analyses of re-admissions that draw
data from the Emory Healthcare CDW.

15. eBIRT
 Integrating “Find an Expert” functionality
based off of existing technologies, such as the
VIVO project
 Kicked off the v2, “Find a Collaborator”
functionality and exploring the different
requirements

16. R-CENTER Web Portal
 Centralizes access to research resources at
Morehouse School of Medicine (MSM)
through the internet.
 Launched in July 2011
 Enables discovery of expertise and resources
for research at MSM, the ACTSI and RCMI
Translational Research Network (RTRN).

17. Aim 2. Engage ACTSI investigators via consultations
 Goal: Maximize the impact of ACTSI investigator
studies and proposals through coordinated use of
bioinformatics, biostatistics, and informatics.
 Carried out in close collaboration with BERD and CIN

18. Aim 2. Engage ACTSI investigators via consultations
 Ad hoc interactions with investigators and research
groups by BIP, CIN, and BERD teams
 Established Studio consultation program
 Investigators request Studio consultation
 a coordinated venue for a pre-review and requirements
evaluation of proposals/projects by a panel of experts to
enhance the impact of ACTSI proposals and projects
 Requests for BIP assistance are captured through the
RAPID system, jointly developed by BIP and the ACTSI
Tracking & Evaluation program

19. Investigator Studios
(a joint operation with BERD, BIP, and CIN)
 Studios started in July of 2010, designed to provide “one-stop
shopping” for pre-submission consultations
 In 2011 there were nine Studios conducted involving the full
complement of BERD, BIP, and CIN faculty
 Ongoing 2012 schedule slots for the first Friday of each month
 Investigators are requested to submit research plans and goals
in advance of the Studio session
 Junior researchers can include their senior faculty mentors in
any session
 Advertising on ACTSI web site and in Weekly Roundup has
been beneficial

20. Aim 3: Informatics Training Program
 Closely coordinated with RETCD
 Clinical Informatics Academy. This Continuing Medical
Education (CME) activity is targeted at clinical researchers,
clinicians, public health researchers, physicians, nurses, and
medical technologists with computer science, engineering, or
biomedical background.
 It focuses on practical aspects of employing biomedical informatics in
research projects and patient care. The course consists of 14 hours of
lecture and breakout sessions.
 The first session was held on June 1-2, 2011 with 42 participants
enrolled. The next course is scheduled for March 2012.

21. Aim 3: Informatics Training Program
 Biomedical informatics (BMI) track in MSCR. A biomedical
informatics track with one student currently enrolled and
another two students to be added in Fall 2011.
 Introduction to Biomedical Informatics is a required course and will
provide an introduction to clinical information systems, bioinformatics,
medical imaging, and computational tools.
 Students will carry out a required translational research rotation and
will take Ethics as another required course.
 Two student slots in the MSCR BMI track will be sponsored with GT
ACTSI BIP matching funds.

22. Aim 3: Informatics Training Program
 Biomedical Informatics PhD Program. In Fall 2010, Emory
obtained approval for a new BMI PhD program that is jointly
administered by Emory’s Departments of Biomedical
Informatics, Math & CS, Biostatistics and Bioinformatics, and
CCI.
 It will engage students with computational and biomedical training within
teams of software system researchers and scientific investigators, addressing
translational bioinformatics and clinical research informatics focus areas.
 Certificate Program in Biomedical Informatics. This program
is targeted at researchers and clinical professionals who would
like to take a set of short courses on the basics and principles
of biomedical informatics.

23. Aim 3: Informatics Training Program
 Clinical and Translational Informatics Rounds (CTIR). CTIR is a
monthly meeting targeted at clinical and translational
researchers, clinicians, pharmacists, nurses and information
services support staff.
 It provides a venue for participants to critically discuss a diverse set of
landmark and current informatics papers, present their work before or after
presentation at national meetings, and brainstorm about current or planned
informatics projects, databases, decision support systems in patient-related
research areas.
 One of the objectives is to form a group of informaticians across the
institution in preparation for the American Medical Informatics Association
efforts to implement subspecialty board certification in Clinical Informatics.

24. Aim 4. Develop novel biomedical informatics
techniques and tools for
 Next Generation Integrative Methods in Medicine. Develop high-performance
computing and data management tools that will make it feasible to systematically
carry out large-scale comparative analyses using high-resolution, high-throughput
datasets.
 Semantic Extract-Transform-Load (ETL) for Data Registries. Develop a semantic
ETL tool that will support temporal concepts and data mappings to semantic
terms.
 Integrity Testing: Biomedical Data Sources and Data Federation. Develop, in a
collaborative effort between GT and Emory, a framework of tools and techniques
designed to detect errors by combining domain knowledge, modeling, and
software testing techniques
 Authentication and Access Control in Federated Environments. Develop a
standards-based security framework in a collaborative effort between GT and
Emory to enhance security capabilities in federated environments.

25. Next Generation Integrative Methods
(in collaboration with Georgia Tech)
 Large volumes of data generated by state-of-the-art
next generation sequencing instruments and image
scanners
 Integration of these data types is limited in research
and healthcare delivery because of challenges with
large scale data management and analysis
 Development of fast methods and tools that take
advantage of
 Large scale storage environments and deep memory
hierarchies
 Clusters of CPU-GPU nodes

26. Semantic ETL Tools and Enhanced Registries
 Linked Databases for Research
 Leverages common data elements and models and
existing standards. Initially for cardiovascular disease,
diabetes and co-morbidities.
 Derived data elements represent categories of data
and temporal patterns of interest.
 Linked to source data – initially, the Emory
Healthcare Clinical Data Warehouse and the Grady
Health System Diabetes Patient Tracking System.
 Supports end-user researcher query and analysis.

27. Federated Security
(in collaboration with Georgia Tech)
 Allow federated management of accounts across
institutional boundaries
 Policy-driven, dynamic authorization based on
attributes.
 Selection of applicable policies and conflict
resolution algorithm occurs in a dynamic fashion.
 Standards based and leverage existing tools:
 XACML, SAML, Shibboleth based standards
 A paper at BIBM 2011 conference

28. Testing of ACTSI Federated Environment
(in collaboration with Georgia Tech)
 Studies involve multiple databases and (complex) data
gathering and management processes
 How to assess the integrity of the federated environment
when Data sources are added, updated, deleted
 A framework to
 Define rules that describe constraints, dependencies,
relationships, and business protocols
 Compose and execute offline and online tests based on
rules and federated databases
 A paper and a poster in AMIA Joint Summit. Another paper
submitted to a software engineering conference

29. Next Generation Exome Sequencing
(in collaboration with MSM)
 Motivated by Minority Health GRID project
 Exome sequencing of specimens from 2400 patients
 Analysis and integration of genomic data with EHR and
observational data
 Development of infrastructure for storage and
management
 High performance computing support through use of
compute clusters

30. Interactions with other Institutions
 The Southeast CTSA Consortium of eight CTSA projects in the
southeastern United States including ACTSI.
 ACTSI BIP leads the informatics group.
 a clinical data sharing initiative to study and develop regulatory
policies, governance and informatics infrastructure surrounding inter-
CTSA clinical and translational research
 Collaboration with the Ohio State University (OSU) CTSA in the
development of a common middleware toolkit to support
data integration and resource federation
 part of OSU-led CTSA Service Oriented Architecture affinity group
efforts
 BIP is pursuing collaborative work with NCBO to integrate
their tools for semantic data modeling into the clinical registry
capability.

31. Interactions with other Institutions
Institution Collaborative Activities
Development of standards-based data sharing framework (initially driven by
University of North Carolina cardiovascular disease research) as part of the Southeast CTSA consortium and use of
BIP’s Analytical Information Warehouse and semantic ETL tools for EHR-linked
bioinformatics and bio-repository infrastructure.
Deployment of Emory eCOI system at University of Florida CTSA. Collaboration on
University of Florida interfacing of eBIRT and VIVO systems.
Collaborations through CTSA Imaging Informatics Working Group and caBIG® Imaging
Mayo Clinic Workspace on informatics tools for secure image data sharing in translational research
and in defining the image data sharing infrastructure in the RSAN image sharing project.
Collaboration through CTSA Imaging Informatics Working Group (IWG) to create common
Duke University Medical Center infrastructure and data models for management and sharing of biomedical image data
and quantitative imaging biomarkers.
Joint design and development of LIMS Study Design Module that has been deployed in
Children's Hospital of Philadelphia both institutions. Shared implementation strategies.

32. Interactions with other Institutions
Institution Collaborative Activities
Collaboration in the CTSA Service Oriented Architecture Affinity Group for development of
University of Michigan interoperable translational research informatics systems. Joint development of integrative
cardiovascular and cancer related research projects. Dr. Saltz serves as Chair of Michigan CTSA
Biomedical Informatics Core external advisory committee.
Collaborations in the Service Oriented Architecture Affinity Group for CTSA, the caGrid
Ohio State University infrastructure development, and the caGrid Knowledge Center -- Emory and Ohio State co-lead
the caGrid Knowledge Center effort. Development of interoperable translational research
informatics infrastructure.
Development of standards-based, federated informatics infrastructure and clinical data
Johns Hopkins University management systems for the CardioVascular Research Grid (CVRG) and application of these
systems in the driving biomedical projects of the CVRG consortium.
Deployment at Emory of REDCap and ResearchMatch systems. Active participation in
Vanderbilt consortium, shared deployment strategies, and extension of REDCap code.