Transcript PPT - University of Connecticut

Biomedical Informatics at UConn & UCHC
Prof. Steven A. Demurjian
Professor, Director of Graduate Studies
Co-Director for Research, Biomedical Informatics Division
Computer Science & Engineering Department
The University of Connecticut
371 Fairfield Road, Box U-255
Storrs, CT 06269-2155
[email protected]
http://www.engr.uconn.edu/~steve
(860) 486 - 4818
1
Biomedical Informatics



Created as a Entity Across both Campuses to Promote
Research Related to:
 Medical and Clinical Informatics
 Bio and Genome Informatics
 Support of Application to NIH’s Clinical and
Translational Science Award Program
Composed of:
 Faculty and Staff Across both Campuses
 New Faculty Hires: 3 in CS&E, 3 at UCHC
Infrastructure Includes:
 HIT Test Bed with EMRs (Vista and Centricity)
 Harvard Profiles Research Software Installation
 Support for CICATS (http://cicats.uchc.edu/)
2
Biomedical Informatics Faculty

UConn Storrs






Steven Demurjian, PhD, Prof. of CS&E & BMI coDirector
S. Rajasekaran, PhD., Prof. and UTC Chair, CS&E
Jinbo Bi, PhD: Assoc. Prof. of CS&E
Maifi Khan, PhD : Asst. Prof. of CS&E
Nasos Bamis, PhD : Asst. Prof. of CS&E
UCHC




Thomas Agresta, MD: Prof. of Family Medicine and
BMI co-Director
Michael Blechner, MD, Assoc. Prof. of Pathology
Xiaoyan Wang, PhD, Asst. Prof. of Family Medicine
Minakshi Tikoo, PhD, Asst. Prof. of Comm. Medicine
3
S. Demurjian BMI Research Interests




Collaborative Extensions to NIST RBAC
 Model When and How Interactions Occur
 Support PCMH and Collaborative Care
Security for XML
 Medical Standards (HL7 CDA, CCR, etc.)
 Customize XML Instances Delivered to Users
Health Information Exchange
 Architectural Solutions for Interoperability
 NIST Prop. on Secure and Privacy Enhanced
Identification
Medication Management & Reconciliation
 Android/IOS Apps Linked to MS Health Vault
 Reconciliation via Harvard’s Smart Platform
 Working with openMhealth.org
4
S. Rajasekaran BMI Research Interests


Data Integration from Multiple Datasets
 Integration of data from multiple datasets is vital in many
areas of BMI.
 For example, there could be multiple records for the same
individual with many providers. Integrating these could
result in great cost savings.
 Prior algorithms can handle only two datasets at a time. We
have developed an efficient algorithm that can integrate
data from any number of sources.
Responsible Data Releases
 Data releases to the public should ensure the privacy of
individuals.
 Several privacy mechanisms, such as anonymization, have
been proposed in the literature. Synthetic datasets are an
example.
 We have introduced a natural notion of privacy and
employed it in the analysis of synthetic datasets.
5
J. Bi BMI Research Efforts

Phenotypical subtyping of complex disease for association studies
Disease phenotypical subtyping
Diagnosis
More accurate
diagnosis
standard
Treatment
Personalized
treatment

subtyping
association
Intelligent online counseling systems to monitor and regulate problematic
human behaviors (eg. college drinking)
OUR SYSTEM
Problematic behavior

Survey
Clinical decision support systems
 Asthma – Computerizing previously paper-based asthma management
program (http://www.labhealthinfo.uconn.edu/EasyBreathing)
 Perinatal care – nursing informatics platform (
http://www.labhealthinfo.uconn.edu/NursingInformaticsPlatform )
6
M. Khan BMI Research Interests



Real-time Monitoring and Preventative Healthcare
 Real-time Access of Patients’ Sensor Data
 Cloud-based Storage Architecture
 Identification of “Early” Symptoms
Reliability and Troubleshooting of Edge Clients
 Troubleshooting Low-power Sensor Devices
 Real-time Failure Diagnosis
Software Architecture for Self-powered Devices
 Fail-safe Energy Management Algorithms
 Energy Harvesting Algorithms
7
A. Bamis BMI Research Interests



Assisted Living and Smart Environments
 Human sensing
 Asset tracking and environmental condition
monitoring using active RFID and Zigbee sensors
 High-level languages and user interfaces for event
detection, and data-driven actuation
Modeling Human Biological Rhythms
 Modeling and detection of deviations in human
biological rhythms
 Android platform for physiological data collection
from wearable Bluetooth sensors
Human Activity Recognition
 Rule-based human activity detection
 Unsupervised detection of deviations and
anomalies in human behavior
8
Thomas Agresta, MD, MBI - Research Interests


Optimal HIT Solutions for Primary Care
 Clinical Decision Support, Patient engagement,
medication reconciliation.
 Health Information Exchange for Care Transitions
 Organizational structures for supporting Primary
Care Informatics
Clinical Informatics Education
 Informatics Educational Methods and Strategies
 Use of Simulation in Primary Care Education
 Virtual patients, families & EMR use in clinical care
 High- Tech, High-Touch Primary Care

Informatics Tools for Collaboration in Clinical
Research Informatics
 Secondary Use of Healthcare data for analysis and
research
9
9
M. Blechner BMI Research Interests

Health Information Exchange
 “Leveraging An HIE Infrastructure To Build A
Clinical Research Data Warehouse”
 CICATS pilot grant for system development to capture
clinical data from a Health Information Exchange
(HIE) into a research data warehouse.


Intelligent Tutoring System Development
 Collaborative e-learning environment leveraging
natural language processing and medical
ontologies (UMLS) to facilitate concept
relationship discovery
Clinical Pathology
 Data warehousing for business and clinical
intelligence in the clinical laboratory
10
10
X. Wang BMI Research Interests

Patient Safety
 Medication safety
 Monitoring adverse drug events from electronic health
records(EHRs)

Fall predication
 i.e. identifying predicators for falls among seniors

Quality control
 Evaluating new and existing therapies in clinical
settings
 i.e. Cost-effective analysis of hepatic resection among
patients with colorectal liver metastasis

Knowledge discovery using EHRs
 Drug repurposing
 Diabetes medications (i.e. Metformin) for cancer
prevention and therapy
11
M. Tikoo BMI Research Interests

Evaluation
 HIE Evaluation studies
 EHR Implementation studies
 Surveys research
 Qualitative research


Performance Metrics and Quality improvement
 Developing middleware to calculate and display
performance measures
 Use of analytics to assist in the implementation of
quality improvement programs
Newly Appointed Position at Dept. Public Health
 Health Information Technology (HIT) Coordinator
for Health Information Exchange (HIE) agreement
between the Dept. of Public Health and the Office
of the National Coordinator
12