1. Knee Rehabilitation Monitor Nicholas Pesce ECE 445 Senior
Design Hall of Fame Spring 2015
2. Introduction There are over 100,000 reconstructive ACL knee
surgeries and 700,000 total knee replacement surgeries annually in
the U.S. Patients are prescribed supervised, physical therapy as a
crucial component for rehabilitation What can be done to reduce a
persons physical therapy time and promote safety?
3. Problems with Knee Rehabilitation Disrupted everyday life
after surgery; dedicated to physical therapy for at least 10 weeks
up to 4 sessions per week Therapists cannot monitor patients at
home activity to ensure safe practice Patients may not be motivated
or confident to perform self exercises without the physical
therapists guidance Therapy is expensive and can cost on average
around $100- $500 per session without insurance There is an
Increased risk of further injury and prolonged rehabilitation when
the patient is unsupervised costing the patient time and money
4. What is Important to Target? How much can a patient flex
their injured knee? How strong are the muscular contractions of the
muscles surrounding the knee? Are patients actually performing at
home exercises? How can therapists see what a patient does at home
to provide better care? How can a patient be assured that the
movements they perform with their knee are safe and cannot reinjure
their knee? How can we give patients the confidence to carry out
their prescribed home exercise routine? How can the cost of
physical therapy after knee surgery be reduced by a significant
factor?
5. The Solution: A Knee Rehabilitation Monitor Features: A 4
flex sensor tracks the amount a patient can flex their knee A
custom Electromyography sensor detects the strength of voltage
contractions in the quadriceps of the injured knee A Texas
Instruments MSP430 Launchpad Microcontroller uses the flex sensor,
an accelerometer and a gyroscope to drive vibration motors which
alert the patient to unhealthy knee movements Data collection
algorithms programmed into the microcontroller collect exercise
data for review by a physical therapist Vibration motors also help
guide patients through prescribed at home physical therapy
exercises in a safe and healthy manner with positive feedback A
battery management system allows the user to detect low power from
the battery source to ensure proper component operation
6. Design Considerations The overall design must be built into
a flexible and wearable sleeve Fabricated microchips must be small
enough to allow easy integration into such a sleeve Fabricated
microchips must be thin and low profile The selected
microcontroller must have sufficient storage capability for logical
operations and programmed algorithms Selected sensors must be
capable of withstanding regular daily use EMG and the motherboard
chip must be separated into two different PCB boards to ensure low
noise acquisition of data Overall design must cost less than the
average cost of a week of physical therapy Power to the chips must
come from a singular 9V battery to ensure small size package
7. System Block Diagram
8. Knee Rehabilitation Monitor Flex Sensor EMG + Flex Sensor
PCB Main PCB + Accelerometer Gyroscope Vibration Motor #2 Vibration
Motor #1 EMG Electrode Push Buttons Front View of Right Leg Side
View of Right Leg
9. Modes of Operation Push buttons on the PCB allow the user to
select between different modes programmed into the microcontroller
Mode 1 = Calibration Mode Used to set benchmarks Mode 2 = Everyday
Mode Used to detect unhealthy movements Mode 3 = Exercise Mode Used
to track and promote at home exercises Mode 4 = Inactivity Mode
Used to prevent knee stiffness
11. Flex Sensor Oriented in front of the knee cap to measure
the knee joint angle Acts as a variable resistor between 10k Ohms
and 35k Ohms, changing its resistance based on how the sensor is
bent Implemented into a voltage divider circuit so that an output
voltage can drive the microcontroller Flex sensor voltage divider
equation Flex sensor
12. Calibration Mode: Flex Sensor Ensures that any patient can
use this device based on their own personal ability A patients
range of motion changes throughout rehabilitation and calibration
accounts for this Captures max/min flexion angles Captures maximum
base quad contraction strength Maximum Knee Extension Maximum Knee
Flexion
13. 0 2 4 6 8 10 12 14 16 18 20 1.5 1.55 1.6 1.65 1.7 1.75 1.8
1.85 1.9 1.95 2 Time (s) Voltage(v) Flex Sensor Data for Heel Slide
Exercise Exercise Mode: Heel Slide Exercise Poor Knee Flexion Poor
Knee Extension Notifies patient if they are coming close to their
maximum flexion and extension during this exercise Give positive
vibrational feedback when the patient surpasses a given
threshold
14. Everyday Mode: Knee Joint Jerk Detection 0 1 2 3 4 5 6 7 8
9 10 1.5 1.6 1.7 1.8 1.9 2 2.1 2.2 Time (s) Voltage(V) Flex Sensor
Data for Jerk Detection 0 1 2 3 4 5 6 7 8 9 10 0 0.05 0.1 0.15 0.2
0.25 0.3 0.35 Time (s) |V|(V) Absolute Derivative of Data (100ms
samples) Jerks Detected
16. EMG Used for feedback during quad contraction exercise
Important for providing a measure of how much the muscles around
the knee have atrophied after repair The signal detected originates
from action potentials fired by cells which are collected at the
surface of the skin with electrodes Based on the relative strength
of a persons quad contraction, more or less voltage will be
detected indicating muscular health/recovery
17. EMG Calculations EMG signals are nested between 20-500 Hz
and are in the magnitude of microvolts-millivolts. Larger muscles
are in the lower half of the band. The resistor and capacitor
values for the EMG band-pass filter were determined to satisfy the
following gain and cut-off frequency requirements: 2 1 10 R Gain R
1 1 1 1 20 2 fc Hz R C 2 2 2 1 300 2 fc Hz R C Band-Pass Gain
Calculation Lower Frequency Cut-Off Calculation Upper Frequency
Cut-Off Calculation R1 = 1k Ohm C1 = 8.5 F R2 = 10k Ohm C2 = 53 nF
Overall Gain = = / Bio-Amp Gain Calculation = (50.5/(120))+1 = 422
V/V
18. Electrode Collection AD622 Instrumentation Amplifier used
to derive a voltage drop (Gain of 422 V/V) LM358 Bandpass Filter
(19.89Hz - 300.29Hz Range) (Gain of 10 V/V) LM 358 Full wave
Rectifiation Detection Filter Full Wave Rectify
19. 0 2 4 6 8 10 12 0 0.5 1 1.5 2 2.5 Time (s) Voltage(v) EMG
Data for Quadricep Activity Exercise Exercise Mode: Quadricep
Activity Strong Contraction Weak Contraction Constant
Contraction
21. Gyroscope and Accelerometer Used to detect motion of the
upper and lower leg Gyroscope: rotational acceleration
Accelerometer: linear acceleration SPI interface Selectable
low-pass filters and full-scale range MPU-6500: Accelerometer and
Gyroscope ITG-3400: Gyroscope
22. Everyday Mode: Twist Detection Compare the output values
between the upper and lower gyroscope abs(upper lower) >
threshold
23. Challenges Single supply op-amp implementation Solved with
voltage charge pumps Data storage Solved with external card
Variations in sensor placement Somewhat solved with calibration
mode Bulky PCB board Could be solved with flexible PCB Technology
Successes Used robust calibration methods to address knee
variations Implemented four feedback algorithms for everyday use
and exercising Created a system that acquires a variety of sensor
data facilitating future feedback algorithms Challenges and
Successes
24. Our goal was to create a product designed for a patients
safety, recovery and to be cost effective If our device reduces PT
time by just 1 week it would be a difference of around $300-$500 in
savings for the patient This product is extremely marketable and
was the winner of the Lextech Most Marketable Product Award which
was one of 3 cash awards given to over 60 competing senior,
Electrical Engineering teams Target Bullseye
25. Add a Bluetooth module and smart phone application to
facilitate data acquisition, to increase processing capabilities
and appeal to users Improve signal analysis for sensor data to
implement more accurate feedback algorithms Allow physical
therapists to easily adjust settings on the device Develop a
smaller PCB board and clean up mechanical design-flexible PCB board
Tech Alternative applications using our technology: Athletic brace
to prevent potential knee injury Weightlifting form correction
Retail or factory inactivity monitor Future Work
26. Acknowledgements My Teammates: Gurmehar Lugani Mark
Hernandez TA Cara Yang Professor Oelze Professor Carney Jonathan
Hernandez, DPT Kristin Buesing, DPT Heather Schaefer, DPT Rebecca
Nef-Heffernan, PT ECE Store and ECE Shop