Transcript Medical database group

Medical Data Analysis/Record Broker
Project Team
Team Members:
Leader: Tracy Giglio: [email protected]
Data Analysis

Kai Unewisse: [email protected]

Galina Walters: [email protected]
Record Broker

Leader: Gundamaraju Lavanya: [email protected]

Aarti Nankani: [email protected]
Jump to first page
Introduction

Customer: Mid America Heart Institute
a
recognized cardiovascular center of
excellence

Current Computing Environment
 does
not support a distributed componentbased computing environment.
Jump to first page
Topics of Discussion








Problems with current environment
Our project(s) role
MAHI database project team’s role
Customer’s role
Functional requirements
Proposed system context
Technology and tools for executing the project(s)
Project Schedule
Jump to first page
Computing Environment Issues

not scalable, flexible, or interoperable
statisticians manipulate data through custom-made
applications/programs wasting valuable time that
could be spent analyzing the data
OCR data is hard to interpret

Duplicate patient_ids




created in the past because the data was not normalized
and database integrity was compromised
requires consolidation consuming human resources
Jump to first page
Our Role - General





Develop OO middleware
Developing a configurable database connection
Extract data from sample database
Perform server side applications
Resulting output is fed to user interface
Jump to first page
Our Role - subproject specific

Data Analysis Broker (project 1)




develop server side statistical applications that
manipulate the data
integrate commercial off-the-shelf (COTS) components
for graphical output.
provide statistical reports in HTML
Record Broker (project 2)



provide work arounds to OCR recognition errors and
database accuracy errors
apply work arounds (algorithms)
end result - return one patient_ID
Jump to first page
MAHI database project team’s role




Provide structure to data (normalize data)
Provide logical validation (consolidate patient records)
Identify relationships
end result - implement relational database
Jump to first page
Customer’s role



Provide user interface
Define requirements
Define project scope
Jump to first page
Medical Data Analysis
Broker
UML Diagrams
Jump to first page
Use Case Diagram

Functional Requirements
Select Multiple Database Columns
Nurse
<<extend>> <<extend>>
MAHI System
Owner/Analyst
Compute Central Tendency (mean,
median, mode) for each column
Compute Frequency Distributions
for Each Column
<<include>>
<<include>>
MAHI Statisticians
Generate HTML Results
<<include>>
<<include>>
Generate Graph Views
Use Case Model: Project 1 - Medical Data Analysis Broker
Jump to first page
System Context Diagram
Jump to first page
Medical Record Broker
UML Diagrams
Jump to first page
Use Case Diagram

Functional Requirements Continued
Enter Patient Attributes
Nurse
Query Patient Attributes
Data Repository
<<extend>>
MANI System
Owner/Analyst
Review Query Results
Create New Patient_ID
<<extend>>
<<extend>>
Run Algorithm to Match Patient
Attributes
Use Patient_ID
<<include>>
<<extend>>
<<extend>>
Review Results of Algorithm
Application
Maually Consolidate Multiple
Record Returns
Use Case Model: Project 2 - Medical Record Broker
Jump to first page
System Context Diagram
Jump to first page
Technology/tools for executing
new development effort

JAVA2 Enterprise Edition 1.3 and 2.0 -provide server-
side and client-side support for developing
enterprise, multitier applications

Together 5.5 -UML modeling tool
JBOSS 2.4.1- application server

ANT 1.4 -Java based build tool

TOMCAT 4.0 / JETTY 3.1.0 - web server



JSPChart - graphics components
MS SQLServer /MS Access 2000 - database
Jump to first page
EJB Architecture
Jump to first page
Project Schedule


See Project Web Site
http://www.cstp.umkc.edu/personal/gwalters/cs551/551_intro.
html
Jump to first page