Transcript Final Presentation - University of Pittsburgh

University of Pittsburgh
Senior Design – BioE 1160/1161
Home ECG Test Kit
James Cook
Carmen Hayes
Joe Konwinski
April 18, 2005
Mentor: Mingui Sun, PhD
Problem Statement
 Heart disease
• 2nd leading cause of death
• Number 1 killer of women
 Limitation of Diagnosis
• In 2003, of the 700,000 Americans that
died, only 148,000 were diagnosed
• Individuals remain unaware of the
symptoms of heart attack or dismiss them
Introduction
 Develop a small device that can:
• Cleanly amplify the electric activity of the
heart
• Save the amplified signal onto a portable
memory solution
 Produce a program that can:
• Analyze the ECG data and give heart-risk
feedback to the user
 Be Sold Over-the-Counter
 Be User Friendly and Safe
Purpose
Purpose is:
• To assess heart risk of seemingly HEALTHY,
middle to upper aged individuals
• To give a cheaper/ less demanding alternative
for healthy patients
Purpose is NOT:
• To replace a health professional in assessing
heart risk of ILL or HEAVY RISK
individuals
Let Us Be Frank….
Market
What’s There…
• Marquette MARS® PC Holter Monitoring &
Review System
--$11,500
• Burdick Vision® Holter Analysis
System
--$6,000
All Marketed Devices are For Use By
Health Care Professionals
Market Size
Based on the 2000 Census, the
population of both sexes ages 40+
was 119,386,252 (42% of
population).
 Family history
 Hypertension (estimated 28 million adults)
 Obesity (estimated 41 million adults)
 Smoking (estimated 20 million adults)
Constraints
 Economic
• Time
• Money
 Regulatory
• False positive
• False negative
• Patient misuse
Project Outline
 Hardware Group
• Develop Miniature ECG Amplifier
• Decide Proper Electrode Placement/Management
• Research Methods on Implementing a Portable
Memory Solution
 Software Group
• Understand the Mechanism behind ECG
diagnoses
• Develop a Computer Algorithm to Interpret
Imported ECG Signal
Hardware Development
Develop a Miniature ECG Amplifier
Hardware Development
Development of Miniature ECG
Amplifier
• Deciding on Chips
• Bread-Board Model
• PCB development
• Testing PCB model
Hardware Development
Deciding on chips
• Size
• Power Consumption
Chip
Quiescent current Supply voltage
INA332
OPA2336
450µA
+2.7 - +5.5 V
20µA
+2.3 - +5.5V
Hardware Development
Breadboard
PCB
Hardware Development
Electrode placement
• Normally has 12 leads,
each one takes a “picture”
of the heart from a different
aspect.
Hardware Development
Normal Recording
Hardware Development
Einthoven’s Triangle
Hardware Development
Goal: Detect P, R, S, and T
waves
Signal
0.4
0.3
R
0.2
T
0.1
P
0
40
40.5
41
41.5
-0.1
-0.2
-0.3
-0.4
S
42
42.5
Hardware Development
Implementing Portable Memory
Current Solution:
• Microchip’s PIC18F2455
• 10-bit A/D Converter
Hardware Development
Hardware Development
Hardware Development
Heart Disease & ECG Timing
Clinically, the electrocardiogram is a
powerful tool in diagnosing certain
types of heart disease.
 Heart block (1st,2nd,3rd degree)
• Timing irregularity in PR interval
 Bundle branch block
• Long QRS interval
 Arrhythmia
• Heart rate too fast or slow (<60; >100 BPM)
 Myocardial Ischemia
• Depressed ST segment
Software Development
Signal Conditioning
 Moving median filter (n=fs) to calculate baseline
Locate areas of muscle contraction using threshold detector
(±0.4V from baseline)
 Find largest gap of continuous data between
contractions for use in further analysis
 If largest gap is smaller than 12 sec,
prompt user to recollect data and to try
to relax his/her body
 Subtract baseline from data
 Use Butterworth 2nd order lowpass
(fl =15 Hz) to remove high frequency
noise
Muscle
Contraction
Software Development
ECG Analysis
 Peak-detector algorithm locates peaks/valleys in the ECG along
with their amplitudes and derivatives
 K-means cluster analysis locates cluster centroids and groups
data points with centroids to minimize sum of squares
 Heuristics based on derivatives,
Q
R
amplitudes, and relative locations
of peaks/valleys give each cluster
T
a label (i.e. P, Q, R, S, T)
•Need to test at least 40
“normal” patients before this
step is complete
P, noise
Q, P, noise
S
Software Development
User Feedback
 Calculate peak-to-peak timing and important slopes of ECG
 Statistical analysis calculates a percent risk of each disease and,
by weighting each disease, a total risk of disease
 User is given feedback consisting of:
Type
Heart Block
Subtype
1st degree
Risk
Weight
92%
2%
10%
5%
0.2%
40%
1st degree heart block is caused by…It results in…It is usually treated by…
nd
2 block
degreeis
1st degree heart
2nd degree heart block…
diagnosed
when the PR interval
degree
is greater than3rd0.2
seconds
3rd degree heart block…
…
TOTAL RISK
11%
Criteria for Success
Completing:
 Design of on-chest ECG hardware
 Ability to detect peak-to-peak time intervals
programmatically
Successes
 Nearing completion of the computer
program
 Developed a miniature ECG amplifier
 Verified that signal peaks can be
detected with chest-mounted electrodes
Future Work
 Finish computer program
 Design and implement flash memory
storage
 Create a working prototype
Acknowledgements
Dr. Mingui Sun
Dr. Marc Simon
Drs. Hal Wrigley and Linda Baker
for their generous donation
Questions?