SeniorProjectSpecifications - Simply Advanced

fullfattruckMobile - Wireless

Dec 10, 2013 (3 years and 7 months ago)

65 views

1

of 8







Specifications

Remote Stand
-
Off Physiological and Vital Signs Monitoring SYSTEM




James Coakley
-

jmcoakle@mail.usf.edu

Danial Goodwin
-

djgoodwi@mail.usf.edu

Chris Mackey
-

cmackey@mail.usf.edu

Yi Zhuo
-

yzhuo@mail.usf.edu




Version 1.0

05
October 2012

2

of 8

Introduction:

The goal of this project is to remotely measure vital signs via a modern smartphone. This document
details the project at a high level.


Key Words:

FFT: This function converting any waveform data into frequency data.

FastICA: Iso
lates mixed signals from each source (RGB in this case) to obtain the best original
signal.

OpenCV: Open Source Computer Vision


Requirements b.ii, b.iv, b.v, c.ii, c.iv, c.v, c.vi




Requirements b.i, b.ix, c.i, c.vii, g.i, g.ii, g.iii, g.iv, g.v


3

of 8

Requirements b.iii, b.vi, b.vii, b.viii, c.iii, c.vii


1.

Body Temperature

Method 1

a.

IR thermometer attached to the phone and connected via USB (or RF transmitter)
pointed at target’s face reads temperature. Temperature reading is sent to the phone
and stored in the Application to be tracked and displayed on the screen.

Method 2

a.

Image of
the body is processed to isolate IR radiation

b.

IR light levels are correlated to known temperatures (filter/compensate for
environmental radiation)

2.

Pulse Rate

a.

Image of the face has blood pumping through it

b.

When blood is pumping through the face the reflect
ivity of certain colors changes

c.

Changing colors form a wave

d.

Wave is processed to isolate blood movement

e.

Peaks in wave are counted to provide pulse rate

3.

Respiratory Rate


Method 1

a.

IR thermometer pointed at the face picks up changes in temperature from
exhaled
breath

b.

Changes in temperature form a wave

c.

Peaks in wave are counted to provide respiratory rate

Method 2

a.

Video of chest is run through motion detection software

b.

Motion forms a wave

c.

Wave is processed with software similar to face recognition to
isolate the number of
breaths

d.

Peaks in wave are counted to provide respiratory rate

4.

Systolic and Diastolic Blood Pressure

Method 1

a.

Analyze the ejection time from the graph obtained in ICA

4

of 8

b.

Obtain the heartrate from the previous subsystem

c.

Plug both data into

the linear equation to obtain each pressure

d.

All other value in the equation are constant therefore heart rate and ejection time
play major role in the change in blood pressure

Method 2

a.

Use camera and microphone simultaneously to capture red peak in face a
nd heart
beat through sound.

b.

Furthermore, analyze the time required for blood to begin travelling at the beginning
of heartbeat to the red intensity peak at the face. This will produce the pressure
value required for measurement.

5.

Blood Oxygen Concentration

a.

Analysis the ratio between red reflection and infrared reflection through video.

b.

The reflection is captured by the video, by analysis the intensity of each reflection at
exact time of each other. We can use this linear equation red/(red
-
infrared) to obtai
n
the O2 saturation percentage.

6.

Facial Gestures

a.

Algorithm takes live video as input and checks for symmetries in the face

b.

Tests for symmetry are tested against a percentage margin in which partial facial
paralysis is not prevalent.

c.

If tests for symmetry fa
ll outside percentage margin, there is an abnormality in facial
functions (ie partial facial paralysis).

d.

If tests for symmetry are within percentage margin, facial function is normal

7.

Pupil Dilation

a.

Algorithm takes live video as input, and takes an initial
frame capture.

b.

After initial frame capture, phone’s flash is enabled and kept on for ~one second.

c.

Right before the end of the one second of flash, another frame is captured.

d.

Second frame is tested against first frame to see if pupil dilation is affected.

8.

E
ye Saccades

a.

Algorithm takes live video as input, and locates eyes and irises. Once eyes and
irises are located, an initial frame capture takes place

b.

After initial frame capture, a second frame is captured, and tested against the first
frame.

c.

This is to be

done with each consecutive frame

d.

If too many frames per second do not match, eye saccades are prevalent.
Otherwise, eyes are functioning normally.

9.

User Interface

a.

See Appendices, Figure 1.

10.

System Program

a.

User uses smartphone to measure vital signs of
patient(s), measurements appear on
screen, then patient sends data optionally.

b.

See Appendices, Figure 2.



5

of 8

Traceability Matrix



Standards

1.

Android Development Tools (ADT) 20.0.3

2.

Android 4.0 API, Level 14

3.

OpenCV, version 2.4.2

4.

FastICA
, Fixed
-
Point algorithm


6

of 8

Appendices



Figure 1
-

User Interface


7

of 8


Figure 2
-

System Program