Transcript Design Poster

Wii System Components to Track Ultrasound Probe and Create 3-D Images
Van Gambrell, Laura Owen, Steven Walston, Jonathan Whitfield
Advisor: Dr. Christopher Lee M.D.
• Amyotrophic Lateral Sclerosis (ALS) is a progressive
neurodegenerative disease that causes atrophy of voluntary
muscles through degradation of motor neurons
• Size and shape of muscles like the bicep are good indicators
of the severity of the disease
• Magnetic Resonance
Imaging (MRI) is the gold
standard for muscle
imaging, but is expensive
and immobile
• Ultrasound (US) imaging
gives comparable 2D
images, is cheaper,
mobile, and has shorter
acquisition times
Figure 1. Diagram of normal and ALS
affected muscle and neural system
APPLICATIONS
• Muscle Volume Analysis
• 3D reconstruction allows for accurate volume calculation
• Provides methods for monitoring disease progression
• Measure muscle atrophy over extended periods of time
• Aid in preventive and therapeutic disease research
• Low cost allows for daily use
• MRI too expensive for research purpose
Figure 2. MRI slices used to
create 3-D muscle image
Figure 3. Ultrasound images used
to approximate 3-D muscle image
DESIGN CRITERIA
Tracking System:
• Provide linear location on a 3-D coordinate system at each
acquisition point
• Provide angular orientation at each acquisition point
• Roll, pitch, and yaw
• Have an acquisition rate that matches the US acquisition
rate
• Have error that is less than or equal to the error from the
US machine
3-D Reconstruction Software:
• Translate slice location to patient location
• Rotate the image according to the angular data
• Recognize muscle tissue
• Compute muscle volume
DESIGN COMPONENTS
•Wii-mote
•Infrared camera capable of tracking 4
infrared (IR) light sources
•Bluetooth data transmission
•Wii-Motion+
•Provides yaw, pitch and roll info
•Arduino Prototype board
•Interfaces between Wii-Motion+ and
MATLAB to acquire yaw, pitch, roll
•Sonosite Titan
•Small portable Ultrasound machine
•2D Linear array transducer
•Design Function
•Two Wii-motes are placed in a known
location and orientation
• Used to calculate the position of
the US probe in 3-D space
•Wii-Motion+ provides rate of change
in yaw, pitch and roll
•Integrated to provide the angular
orientation of the ultrasound probe
•IR LEDs were tested as a redundant
source of yaw, pitch and roll angular
data
•Each acquired US image is assigned a
x, y, and z position and a yaw, pitch and
roll angle from the Wii-motes and WiiMotion+
•3-D Reconstruction Software
•Takes position data and realigns slices
in new matrix representing scanned
area
•Uses angle data to rotate slice
appropriately
•Done through post processing
DESIGN PERFORMANCE AND SUCCESS
Figure 4. The Arduino prototype board
Figure 9. Position and angle data for each acquisition slice for both
systems. LED System (Left). LED + Wii MotionPlus System (Right)
FUTURE DIRECTIONS
Figure 6. LED+WiiMotionPlus
System with Sonosite Titan
SYSTEM VERIFICATION
•Tested by moving source on a fixed path
with known distance
•Compared measurements with
returned data
•Created a phantom for volume accuracy
•To compare US and MRI
•Linear error is greater than that of the
ultrasound system
•US system error is within 3 percent
compared to 22, 31, and 2 percent
•System failed accuracy verification
•External IR interference
negatively affected ability to
accurately track LEDs
•Reconstruction Software:
•Resolution reduced from
640x480 to 160x120 in order
to function within memory
constraints of MATLAB
•Scans limited to less than 15sec
due to Wii-Motion+ sensor drift Figure 8. Drift from Wii Motion+ with
• Angle and position data filtered movement only from 17-27 seconds
due to significant noise
Figure 5. Arduino with LED + Wii
MotionPlus attached to US probe
3D position Error
35
• Construct a stable attachment device to connect the position
sensors and the US probe
• Design a more efficient infrared LED arrangement for position
tracking of US probe
• Explore different methods of obtaining x, y and z position of US
probe to increase accuracy
• Use of other accelerometers and gyroscopes
• Streamline software for a more user friendly data analysis
• Determine methods to reduce effects of sensor drift
• Devise a more accurate scaling for Wii-Motion+ output
30
25
Percent Error
INTRODUCTION
ACKNOWLEDGEMENTS
20
15
10
5
0
x Error
y Error
z Error
Figure 7. Linear error from LED system
Special Thanks to our advisor Dr. Christopher Lee M.D. of the Vanderbilt
Department of Neurology, Dr. Paul King, graduate student advisor Alex
Makowski, VUMC for the use of a Sonosite Titan, Nintendo, Jordan Brindza
and Jessica Szweda of Notre Dame for the WiiLAB software package and
manuals, and Knuckles904 for help with the Wii-Motion+ Arduino code.