School Of Engineering

ranchocucamongabrrrAI and Robotics

Nov 6, 2013 (3 years and 11 months ago)

70 views

Page:
1
/
10


School Of Engineering



E
NGR

402
:

Senior Design Studio Seminar

Assignment:
Project Proposal

Course Instructor:
Prof. Robert Klenke

Due Date:
Sep

30
th

2013

S
tudent Name: Shah, Gaurav M

& Francy, Scott.


Page:
2
/
10


Contents

Executive Summary

................................
................................
................................
..................

3

Introduction and Background

................................
................................
................................
...

3

Project Proposal

................................
................................
................................
........................

6

Project Timing

................................
................................
................................
..........................

7

Project Considerations

................................
................................
................................
..............

9

Economic Considerations

................................
................................
................................
......

9

Ethical Considerations
................................
................................
................................
...........

9

Manufacturability

................................
................................
................................
..................

9

Health and Safety Consideratio
ns

................................
................................
.......................

10






Page:
3
/
10


Executive Summary


All medical facilities have monitoring systems to raise awareness of problems for timely
response to patients whose health can
change dramatically in a short period of time.
Equipment that monitors vital signs is important, but there are other clues and gestures
that can be captured through simple patient observation. Through automated video
monitoring with event detection, a pa
tient’s motions could be captured and stored as
events. Hand movement, eye blinking, sleep time and other events could be individually
detected and recorded. This system would be especially useful for monitoring patients in
critical condition or coma pat
ients. A second application also exists for monitoring
infants. This system could be used to study a large sample of infants by capturing events
and recording them. These results could then be used to establish milestones for progress
and development of

a newborn relative to a statistical average or expectation.

This system will function by passing camera video through an image processing board.
A server application will interface with this board to record detected events into a
database and also record

video. The server application will also serve a client application
for users to view and analyze recorded events and video.

This project was chosen because of the potential benefit it could provide the medical
community by its application. The desired l
ong term goal is for this system to be an early
detection system for illnesses and infant development to identify problems timely and
enable healthcare providers to treat them as early as possible.

Introduction and Background

Event monitoring is not a new
concept to the medical industry. Nearly all medical
facilities are equipped with monitoring equipment today to capture a patient’s blood
pressure, heart rate and other vital signs. In recent years, new, more mobile systems have
been developed to monitor
a patient’s vital signs and other parameters such as
environmental conditions and patient activity.

A literature search yielded a similar project to monitor a patient’s vital signs, capture and
store events, and provide data to medical personnel through a
client application. A group
at the University of Alabama conducted this project with title “System Architecture of a
Page:
4
/
10


Wireless Body Area Sensor Network for Ubiquitous Health Monitoring”. The project
proposed herein differs in both the focus as well as the

method. The system described in
this proposal will focus on movement or activity monitoring instead of a comprehensive
monitoring system. Rather than capturing heart rate, blood pressure, temperature and
other variables, this system will focus on detect
ing and categorizing activity. This system
also differs in method, compared to the University of Alabama project. The project team
at the University used various sensors attached to the body to record the desired data
from the patient (
Figure 1)
. This m
ethod is both intrusive and obtrusive and can be
uncomfortable for the patient. Video will be used in this application, which eliminates
the problem with patient comfort. It is also more portable and potentially easier to deploy
depending on configuratio
n of the system in certain environments.

Figure 1


A US patent also exists (6406426) for “Medical Monitoring and Alert System for Use
With Therapeutic Devices” which precedes the University of Alabama project. This
system is intended to be a monitoring s
ystem for patients in a medical facility for alerting
medical personnel where the University of Alabama project could be used to monitor
patients’ health remotely. The system proposed herein is designed to be used in a
medical facility for both research a
nd diagnostic information rather than a real
-
time
Page:
5
/
10


alerting system. The system described in the patent does not use video; it uses sensors
that communicate wirelessly and attach to the body.


Figure 2


Video event detection systems are also common today b
ut the application is typically for
security. Event detection in security systems typically detects movement in a broad
frame of video and uses the event as a trigger to record video. This reduces the amount
of video that needs to be recorded and saved,
thus reducing cost for storage. The system
for patient monitoring will be different because it will identify specific movements in
video and categorize them as specific events rather than simply detecting any movement
in frame and declaring an event.

To c
onclude the summary of research findings, although similar technology exists, no
system is known to exist for the application described in this proposal implemented. The
application, intended use and method of operation are unique.

Page:
6
/
10


Project Proposal

This p
roject proposes to design a camera / video based monitoring system to capture and
identify events defined as patient or subject movements. The design concept is to
develop a circuit for capturing video from a camera and output this to the Xilinx Virtex
-
4
mini module. Through a PC interface to the Xilinx mini module, the video signal will be
recorded by the server application at a user specified frame rate. The video signal will
also be analyzed in hardware to determine differences between the last record
ed frame to
a previous frame of difference N to be determined. An algorithm will be developed that
will determine if the frame has changed more than some predetermined threshold
compared to the previous frame. The threshold will be controllable through a

sensitivity
setting that can be adjusted in hardware.

When the threshold is met for difference between the 2 images, the event trigger will be
enabled and the two images will be sent to the server application for further processing.
The server applicatio
n will determine where the change has occurred and categorize the
movement. That event and associated details such as timestamp, event type and patient
ID will be stored in the database.

The server will also listen for connections from the client applicat
ion. When requested
by the client, the server will provide data from the database as well as video that had been
recorded previously. All components will communicate using Ethernet.

Figure 3

For this design, a final decision needs to be made about the handling of image processing.
The options exist to process the image in either hardware or software. It is believed that
hardware processing would improve image
-
processing performance but may b
e more
difficult to implement compared to software. Regardless, the detection algorithm will
need to be written to identify events properly. The video signal will need to be analyzed
Camera


Capture patients

Board


read camera(s)


process image for
server


communicate with
server

Server


process image


detect events and
store in DB


record video


Serve the client

Client


Display results


can be multiple device
/ plateform

Page:
7
/
10


and a determination made how best to calculate differences. The Xilinx

interface to the
server application will also need to be developed to determine how best to capture video,
receive an event, and perform further image processing if necessary. The steps required
for the planning and execution have been broken down into m
ajor tasks and will be
considered milestones when the task is completed successfully. The tasks are described
next and also summarized in
Table 1.

The first five milestones are all related to image capture and processing. The first task
would be to analy
ze the various options for the video interface and select the best
interface option. Wireless Ethernet would be the preferred interface option, but a
determination of how the raw video signal is captured and analyzed needs to be made.
The signal needs to

be sampled at a certain rate and the samples need to be
mathematically compared to determine an event has occurred.

The seventh milestone involves interfacing the hardware to the PC where the server
application will record the video signal and store event
s and associated data in the
database when a trigger is sent. The other milestones in the design phase involve
planning the architecture and interfaces for a server application that manages storing the
video data, performing additional image calculation a
nd accepting client connections to
view this data.

The next set of milestones is for project execution where the design will be implemented
with intermediate testing at each step. The hardware should be procured no later than
week 3 of the Spring semester but can actually be procured after the design is f
inalized
and the design report is submitted.
Project Timing

Project work will occur in both independent team member work sessions as well as
collaborative sessions. A weekly team meeting will be held and each will have three
objectives: to review status

of the prior week’s activities, discuss any problems or
barriers, and set the expectation and activities for the following week, updating the
schedule if necessary. All the planning will occur during these meetings and will be
tracked using Microsoft Pro
ject or equivalent software. Milestones will be entered into a
shared calendar that team members have access to and can view at any time.




Table 1

Phase

Description

Week

Responsible

Design

Submit Project Proposal

5
-
Fall

Francy

Design

Analyze video
interface options, select interface

7
-
Fall

Shah

Design

Simulation of video capture circuit and Xilinx interface

8
-
Fall

Francy

Design

Analyze video signal, determine sampling
, signal processing

9
-
Fall

Francy

Design

Image calculations and algorithm for
image differences

10
-
Fall

Shah

Design

Develop event detection method(s)

11
-
Fall

Shah

Design

Design interface between PC and Xilinx mini module

12
-
Fall

Francy

Design

Design server application and database schema

13
-
Fall

Shah

Design

Design and plan
client application and interface

14
-
Fall

Shah

Design

Submit Design Report

2
-
Spring

Francy

Execution

Procure hardware and parts

3
-
Spring

Shah

Execution

Build capture circuit

4
-
Spring

Francy

Execution

Build interface to Xilinx mini module

7
-
Spring

Shah

Execution

Code event detection algorithm

9
-
Spring

Francy

Execution

Code server application and test database

10
-
Spring

Shah

Execution

Code client application and test functionality

10
-
Spring

Francy

Testing

Complete first round of performance testing

11
-
Spring

Francy/Shah

Testing

Refine image processing and/or detection algorithm

12
-
Spring

Shah

Testing

Complete second round of performance testing

13
-
Spring

Francy/Shah

Demo

Senior Design Expo

14
-
Spring

Francy/Shah

Report

Final Report

14
-
Spring

Francy


The technical skills required to execute this project include signal processing, circuit
design, systems design, C programming, Ethernet and networking. Both team members
have skills and experience in these areas so assignments can be changed later if it
is
determined one team member has stronger skills at a particular task compared to the other
team member. Initially, the tasks will be assigned an owner as indicated in
Table 1
.




Project Considerations


Economic Considerations

A small budget exists for th
is project to produce a system suitable for demonstration.
The budget is a rigid boundary as limited funding is made available by the University for
the project. Additional funding could be requested from the project sponsor if necessary
but will need to

be justified accordingly. It is not believed at this stage that the budget
constraint will have a major impact on the project because the components currently
being investigated for use in the system should fall within the budget. If it is determined
th
at these components are not adequate or additional purchasing is required, then the
budget limit would be met.


Ethical Considerations

Because of confidentiality between a physician and patient, it is believed that data
recorded and stored by this system c
an be treated like other patient data when used for
analysis and comparison to a previously identified statistical normal or expectation for a
given study. When used for research over a large sample though, the data must be kept
anonymous and shared witho
ut releasing confidential patient data. The proper use and
handling of this data is the responsibility of medical personnel and cannot be addressed
with a technical solution.


Manufacturability

This project will both integrate existing technologies and ma
nufacturing a circuit used
with the video capture. The circuit would include common electrical components that are
easily obtained. A PCB would need to be designed and manufactured for this component
but this should also be easy to obtain given the amoun
t of PCB manufacturers available.
A consideration for the components used for this project is if the system were to be
implemented on a large scale, the selected parts would need to be available and the
design would need to be able to interface with other

similar equipment in the event a


component is no longer manufactured. Lack of control over supply chain for some of the
included components is a drawback compared to designing and building all the
components used within. For this reason, alternative com
ponents will be selected in case
of supply problems.


Health and Safety Considerations

In the design of this system, health and safety considerations were made regarding how
the system would be installed. Wireless communication is desired to eliminate
wires
being an obstacle or hazard while the system is in use. It is possible that wireless
communication is not available in certain environments so a wired option was also
considered. This would make installation more complex as well as add complexity
m
anaging two different versions of the system. It is understood that the system may be
restricted to certain environments where wireless communication is possible but wireless
communication can also be included with the system but self
-
contained making it
a
customer responsibility to interface to it from another network.