Biofeedback Analyzer Project

1
Biofeedback Analyzer Project

• Project Team Ben Becker
• Kurk Laird
• Jim Tabbert

2
Main Purpose

• Real-Time Observation of
• Heart Rate
• Skin Temperature
• Muscle Tension

3
What For?

• An analytical tool to determine effectiveness of
  treatments
• An analytical tool to determine effects on muscle
  groups
• Active Biofeedback Therapy

4
Our Goal

• Current production units cost gt 4000
• Our goal was to build a unit for under 600
• To cut major expenses the unit had to be a
  peripheral instead of stand-alone

5
Key Deliverables

• Software
• Host Program with GUI
• FPGA Program
• Hardware
• Electromyography (EMG)
• Heart Rate
• Temperature

6
Timeline

• Software
• Immediately deviated from the timeline due to a 3
  week delay in shipment of our USB development
  board
• Lack of knowledge on the material slowed progress
  significantly
• Unable to implement a database
• Hardware
• Misjudged the appropriate duration of key tasks
• Due to limited initial knowledge,
  incompatibility, and nonfunctioning sensor
  designs
• Unable to interface the EMG and Heart Rate sensor
  with software

7
Resources

• The project team spent approximately 330 hours
  over the 5 month span.
• The team budgeted 600 for the project and spent
  approximately 395.
• 65 budget usage.

8
Risk Mitigation

• Problems were dealt with as they arose
• Solutions were discussed by
• project group
• project advisor
• professors
• Problems were thoroughly documented in the event
  log and bi-weekly status reports.

9
Software

• The software is broken up into 2 sections
• Host Program
• FPGA Program

10
Host Program

• MFC Application written in C
• Written using Microsoft's Visual C .net
  development environment.
• User interface for the entire project
• All control functions are handled by the Host
  Program

11
Host Program
12
FPGA Program

• Written in VHDL using Altera Quartus II
  development environment.
• Handles data gathering and transmission.

13
Software Roadblocks

• Host Program
• Interfacing the host program with the FTD2XX
  driver
• FPGA Program
• No initial knowledge of VHDL

14
Electromyography

• What is Electromyography (EMG)?
• Electrical potentials created by the movement of
  the muscle.
• How does it work?
• When the muscle contracts it gives off electrical
  signals in the range of 10-500 Hz.
• Amplitude 1 10 mV rms

15
Electromyography

• How is it detected?
• Bipolar electrode
• Differential Amplifier
• Feedback

16
Electromyography

• Differential Amplifier stage (Gain 1000)
• Filter (12Hz cutoff)
• Amplifier w/ adjustable gain (1-21)
• DC bias stage

Differential Amplifier w/ Feedback
Filter
Amplifier
DC bias
17
Electromyography

• Our EMG

• Commercial EMG Unit

DelSys
18
EMG Roadblocks

• Interfacing the Electromyography sensor with the
  USB development board via analog-to-digital
  conversion

19
Heart Rate

• How does it work?
• Infrared phototransistor pair (Honeywell HLC1395)
• Detects small variations in skin contrast
• 80 - 100 mV AC sine wave
• .9 Hz to 3.5 Hz (50 200 bpm)

20
Heart Rate Sensor

• Architecture
• Amplifier (Gain 20)
• Bandpass filter
• Lowpass filter

21
Heart Rate Roadblocks

• Delayed shipment of replacement HLC1395
• Consistent operation of the HLC1395.
• Properly filtering the device to eliminate
  unwanted frequencies

22
Temperature Sensor

• Original Design - not used
• Abandoned for high precision, linear, skin
  contact circuit
• Sensor Used
• YSI 709B probe
• YSI 44302 resistor set

23
Temp. A/D Conversion

• Not as simple as expected
• ICL7109 12-Bit Parallel Converter
• Suggested by Professor Most
• Provides continuous output
• Direct Parallel connection to USB Development
  Board

24
Summation

• Project Goal
• Self-contained, USB Compatible Device Created
• 2 of 3 sensors operational
• Host and FPGA programs functional
• Problematic Deliverables
• A/D of EMG and Database
• Value to Society
• Significantly reduced cost alternative
• Real world demand

25
Summation

• Project Highlights
• Extensive knowledge gained
• C and VHDL Coding
• FPGA Usage
• Sensor development
• Conversion Experience
• Resources and Costs
• 395 spent
• 330 Manhours
• Demonstrations during Poster Sessions

26
Appreciation

• Special Thanks
• Project Advisor Professor Warren Klope
• Technical Advisors
• Professors Robert Most, Clare Cook, and William
  Liew
• YSI Temperature, Incorporated
• Brent Eresman
• General Thanks
• EEET/CNS Dept.
• Ferris State University

27
Questions?