|
Current Research
Gesture Recognition and Surgical
Skill Evaluation
Johns Hopkins University,
Computational Interaction
and Robotics Laboratory (CIRL)
Professor Gregory Hager
January 2007-Present
Previous Research
Haptics
for Robot Assisted Minimally Invasive Surgery
Johns Hopkins University,
Haptics Exploration Laboratory
Professor Allison
Okamura
August 2004-December 2006
Haptic
Integration of IBM Manipulator
Santa Clara University, Robotics
Systems Laboratory (RSL)
Professor
Christopher Kitts
September 2003-May 2004
Underwater Robotics
Santa Clara University, Robotics
Systems Laboratory (RSL)
Professor
Christopher Kitts
September 2001-May 2003
Parallel Computing
University of Delaware, HIPERSPACE
Professor Lori Pollock
June 2003-August 2003
~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~
CURRENT RESEARCH
I am conducting research on
surgical skill evaluation project. Can we create a motion
vocabulary analogous to phonemes in speech recognition to
objectively evaluate surgical performance? I am looking at three
tasks: suturing, knot tying, needle passing to determine the
vocabulary and motion while using a robotic MIS system. Evaluating
vision and robotic kinematics, dimensionality reduction is used and
features are extracted to build good classifiers.
I briefly worked with Balazs
Vagvolgyi by recording surgeries and helping make this video on his
project of 3D augmented reality displays. The overlays are
constructed by registering tumors based on CT scans for laparoscopic
partial nephrectomy.
~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~
MASTERS RESEARCH
Haptics for Robot-Assisted Minimally Invasive Surgery
[ Sensory Substitution for Medical Devices | Sensory Substitution vs. Direct Haptic Feedback]
For a crash course in my research, click
HERE to see the invited talk I gave at Santa Clara University.
Sensory Substitution for Medical Robotics
With the high costs in the OR,
creating surgical systems that integrate novel computer and
human/machine interface technologies will revolutionize surgical
procedures, extending the surgeon's abilities to achieve better
outcomes at lower costs. Haptic (force and tactile feedback) has
been proposed as a way to further enhance the performance of these
systems. A limitation to the current generation of MIS robots is the
lack of haptic feedback where the operator relies solely on visual
feedback to determine the amount of force being applied. Implementing
direct haptic feedback to the surgeon's hands remains impractical for
clinical application because of the cost and time of applying force
sensors to disposable tools and the current limitations in sensing and
control technologies. The goals of this work are to develop an
intuitive augmented reality system for feedback of force information
through sensory substitution, and to evaluate its performance in a
surgical task such as knot tying.
Work done in collaboration with graduate students
Tope Akinibiyi.
The video shows the difference
with and without sensory substitution. The first one is the robotic arms being controlled by
the surgeon tying an instrument tie. The second is a 2D
representation of what the surgeon sees in his/her console
with force sensors added onto the instrument.
Funding Source: NIH Grant R01 EB002004, Whitaker R6-02-911
Special Thanks to the Minimally Invasive Surgery Training Center (MISTC)
and Dr. David Yuh. ~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~
UNDERGRADUATE RESEARCH
[
Haptics | Underwater Robotics
| Parallel Computing |
Sound Manipulation]
Haptic Integration of IBM Manipulator
[PDF
of Senior Design Report]
As the first design team at Santa Clara
University to experiment with force feedback and integrate haptic
feedback into a robotic arm, helping make tasks more realistic for
robotic operators. The main objective of our interdisciplinary
team of six was to modify an existing IBM robotic arm, model number
7545, focusing on the integration of haptics into the operation and
control of the robot. This integration was to allow the user to feel
differences in the rigidity of the objects picked up by the robot
gripper. Therefore, the system would be providing cutaneous feedback
to the user, so that the user will be able to respond to forces. For
example, if the gripper was to pick up a rock, the user could
distinguish the hardness of that object as opposed to squeezing a
spring, which would feel pliable. Another objective was to perform the
proper maintenance required for the robot since it was not functional
and out of use. After getting the robot in a functional state, a new
user interface was designed and manufactured that would not only move
the arm and gripper, but also be simplistic and user friendly.
The video shows picking up and
egg with and without haptic feedback.
 
The pictures of user end gripper (left),
Senior Design Team - Force Feelin'
(center), a picture of the IBM
manipulator with robotic gripper (right)
Program Slicing for OpenMP
Shared Memory Parallel Programs
- Summer 03
I spent ten weeks in the
summer after my junior year doing research on parallel computing at
the University of Delaware as part of the NSF CRA-W
program. After studying the current program
slicing algorithm, I designed and implemented a user friendly slicing
tool for OpenMP written
by Matthew Bridges.
This privilege gave me the opportunity to experience the day to day
life of a Ph.D student.
The interface was made so that a parallel programmer who knows little
about the underlying infrastructure could run it. Slicing is used for
software development and maintenance activities such as program
understanding, software testing, and debugging. By extending and
modifying an interprocedural slicing algorithm for sequential
programs, and an intraprocedural slicing algorithm for parallel
programs, we have developed a technique for static interprocedural
slicing of shared memory parallel programs, written using OpenMP
explicitly parallel constructs. OpenMP is the standard for explicitly
parallel shared memory programming. To see a journal of my experience
that summer, click
HERE
Underwater Robotics - Fall 01->Spring 03
During my sophomore year, I built a rudimentary,
low-cost tethered vehicle is made of PVC tubing and uses bilge pumps
as thrusters along with six other students. Our
interdisciplinary team installed a camera and an internet control
system allow students to fully control the vehicle and to rapidly
prototype new ideas on a simple engineering platform. Throughout the year, we
constructed a low-cost but robust Remotely Operating Vehicle which
consisted of PVC pipes as a frame, six thrusters for
propulsion and maneuvering, a tether which supplies controlled power
from a 12 volt DC power source, floats, and a controller using push
buttons to control the thrusters. The robot sent a stream of video so
that a recording was displayed on deck. I configured and tested a safety system for the
robot. This involves programming a BasicStamp microcontroller,
interfacing relays and setting up communications between the basic
stamp and DSP.
Picture of Seafox (courtesy of SCU RSL)
That summer, I worked independently with Professor Christopher
Kitts to do research and expand the vehicle’s capabilities. Utilizing my electronics experience, soldering and finishing
up the wiring and circuitry, I used my knowledge of C coding and
programmed the underwater ROV to be controlled through the Internet,
through the relays, through the Z world microcontroller, and through
the hand held controller.
Junior year I progressed to work
on the Santa Clara University’s Triton undersea robot during my junior
year. This shallow-water (<1000 ft) 3 Hp
tethered vehicle is being used for a variety of marine science studies
as well as for several robotic technology studies.
I became in charge of coding. I
analyzed and tested the capacity of the code and then worked to
manipulate the abilities of the robot.
Pictures of Triton (Courtesy of SCU RSL)
Sound manipulation and web development with Prof. Ogunfunmi -
Fall 00->Spring01
Starting fall quarter freshman year, I took the initiative of doing
research learning C++ and the principles of digital signal
processing. My project expanding to working with web-based
applications and manipulating sound clips.
|
