1. Roy Jung, SangChul Shin, XinCheng Jin
daeyoung@eden.rutgers.edu, sangchul@eden.rutgers.edu, jamiejin@eden.rutgers.edu
Advisor: Prof. P. I. Reyes , and Prof. Y. Lu
ZnO Thin Film Resonator – Based Heart Monitor
Introduction: ZnO Films And
Nanostructures For Sensing
• ZnO is Multifunctional: piezoelectric,
semiconducting, optical, transparent
conductive
Advantages
•Small dimensions
•Controllable operating
frequency by varying Mg
composition
•Low power consumption
•Can be integrated with
Si-based electronics
•Wireless frequency
range
Conclusion
TFBAR Characterization
• Directly applied voltage to ZnO biosensor
contact region to obtain its bandwidth region
• Bandwidth regions
• 411.02MHz to 420.47MHz
• 1.2426GHz to 1.2855GHz
We combined the ZnO nanostructure-based
biosensors with cardiac arrest symptoms
detecting software to create a complete system.
This is an easy to use, low power consuming
device. Since it is Si based, it can be offered to
anyone at an affordable price. This will help to
lower number of heart related deaths because
of its usability and price.
Bio-Sensor Quality Test
3 different inputs
• Sudden Pressure, Sustained Pressure, and
Repeated taps at 2Hz.
Apply 3 different
inputs
Record data
through an
oscilloscope
Create a filter that
can reduce signal
noise on MATLAB
Find Q, R, S Points
Calculate average
Rise time and
Recovery time
Calculate total
reaction time
Input Noise Filtering
• FIR1 filter with order 25 and rectangular
window outputted best noise cancellation
result.
Filtering Result
Application of
FIR1 filter
Q,R,S Detection
Reaction time
• Q point is a minimum voltage location
before R point.
• S point is a minimum voltage location
after R point.
• R point is a maximum voltage location.
Q,R,S Detection Result
• average rise time is 0.027568 seconds
• average recovery time is 0.00363 seconds
• average total reaction time is 0.031198
seconds
Data Collection
• Data from http://www.physionet.org/
• Data contains detailed analysis about
patients’ information, number of isolated
beats, noise, Arrhythmia type.
• 150 ECG data from Arrhythmia patients and
healthy test subjects.
• Data was categorized based on patient’s age,
gender, and number of isolated beats.
Human Pulse Detection
• Average Beat-to-Beat Interval (R-R Interval) is 0.9468 sec
• Pulse Rate is 1.0562 Hz
• Experimental Subject Condition: Healthy
• Average Beat-to-Beat Interval (R-R Interval) is 1.3324 sec
• Pulse Rate is 0.7505 Hz
• Experimental Subject Condition: Healthy
• Step 1: Take the mean value of the R-R intervals in the
normal condition R-Raverage
Actual Patient Data During
Normal Condition
Actual Patient Data During
Tachycardia
• Step 2: Take the standard deviation of the R-R
intervals in the normal condition R-RSTD-Normal
• Step 3: Take the standard deviation of the R-R
intervals in a moving time window R-RSTD-Test(t)
• Compare:
 If R-RSTD-Test(t) < or = R-RSTD-Normal patient is in
normal condition
 If R-RSTD-Test(t) > R-RSTD-Normal patient is
experiencing arrhythmia
Sauerbrey's formula based pressure variance
c66 is the stiffness constant of the piezoelectric material, and ρ is the density of
the piezoelectric material. c66 = 4.43 X 1010 N/m2 and ρ = 5680 kg/m3
• Placed the bio-sensor under a wristband and
connected to the oscilloscope.
• External pressure on the bio-sensor created
noise.
Human Pulse Noise Filtering
• Convolved input signal with 5Hz Low Pass Filter
Theoretical Algorithm
Heart Attack Detection Software
Motivations And Goals
• Arrhythmia Detection
𝑇
𝑁𝑜𝑟𝑚𝑎𝑙 0 𝑖𝑓
𝑖=1
𝑁
𝑦𝑖
𝑁
≤ 𝑡ℎ𝑟𝑒𝑠ℎ𝑜𝑙𝑑
𝐴𝑟𝑟ℎ𝑦𝑡ℎ𝑚𝑖𝑐 1 𝑖𝑓
𝑖=1
𝑁
𝑦𝑖
𝑁
> 𝑡ℎ𝑟𝑒𝑠ℎ𝑜𝑙𝑑
• Yi = Input signal
• Threshold value is set during synchronization
process
• To create an easily accessible and
affordable heart monitoring bracelet for fatal
heart malfunction detections.
• Technological goal will consists of systemizing
database library for heart malfunction, creating
an algorithm for a potential heart malfunction
and creating a threshold variable for system
application.
References:
1. “ZnO Nanostructure-Modified QCM for Dynamic Monitoring of Cell Adhesion and Proliferation”, to appear in Biosensors and Bioelectronics”, 2012. (Pavel I. Reyes, Ziqing Duan, Yicheng Lu, Dimitriy Khavulya and Nada
Boustany)
2. Galoppini, Elena. "Multifunctional ZnO-Based Thin-Film Bulk Acoustic Resonator for Biosensors." Journal of Electronic Materials (): 1605-1611.
3. Massachusetts Institute of Technology. (24 May 1997). MIT-BIH Arrhythmia Database Directory (Hypertext edition).Available: http://www.physionet.org/physiobank/database/html/mitdbdir/mitdbdir.htm
4. Goldberger AL, Amaral LAN, Glass L, Hausdorff JM, Ivanov PCh, Mark RG, Mietus JE, Moody GB, Peng C-K, Stanley HE. PhysioBank, PhysioToolkit, and PhysioNet: Components of a New Research Resource for Complex
Physiologic Signals. Circulation 101(23):e215-e220 [Circulation Electronic Pages; http://circ.ahajournals.org/cgi/content/full/101/23/e215]; 2000 (June 13).
5. Amro,(20 November 2009). MATLAB: filter noisy EKG signal. Available: http://stackoverflow.com/questions/1773542/matlab-filter-noisy-ekg-signal
6. Kunt, M, Ligtenberg, A , A robust-digital QRS-detection algorithm for arrhythmia monitoring , Computers and biomedical research , vol 16 , no 3, p.273 – 286
7. Tsipouras, M, Fotiadis, D, Automatic arrhythmia detection based on time and time-frequency analysis of heart rate variability, Computer Methods and Programs in Biomedicine, vol 74, p.95 – 108
8. Gothwal, H, Kedawat, S, Kumar, R, Cardiac arrhythmias detection in an ECG beat signal using fast fourier transform and artificial neural network, J. Biomedical Science and Engineering, vol 4, 289-296
Frequency(GHz)
dB(A.U.)
Patient Type Mean R-R Standard Deviation R-R
Female Patient with
Arrhythmia
0.6381 sec 0.2301 sec
Male Patient with
Arrhythmia
0.6629 sec 0.3154 sec