Background and Motivation
VR
technology has advanced significantly in the past decade. Technology
for haptic interfaces has advanced to such an extent that the
market has segmented into two groups: high-end devices costing
over US$10,000, and low cost devices averaging approximately US$100.
The high-end devices are developed for, amongst other areas, haptic
research and surgical simulation, while the low-end devices are
produced mainly for the PC video game industry, and more recently
are appearing as OEM components of complex interfaces, such as
automobile cockpit controls.
This
price gap can be explained partly by the disparity in production
size and complexity among categories. A significant price reduction
is achieved by only generating simple force profiles. This makes
it possible to build all controller hardware into the device itself,
and then only maintain low bandwidth communication between the
device and the computer. Differences in design objectives are
another cause of the price gap. Gaming interfaces are designed
to merely convey the sense of a force event or texture, thereby
allowing the use of inexpensive components. However, the design
objective in interfaces for laparoscopic surgery simulation is
to render forces accurately enough to match or exceed the limitations
of human force perception. Whether such high standards for laparoscopic
surgery simulation are actually needed has not yet been demonstrated.
Arguments can be made to the contrary. Research in multimodal
perception shows that in many contexts we rely more on vision
than touch to judge quantities such as size, shape and position.
In judging stiffness in the face of contradictory cues, the perceptual
system may discard haptic information altogether. Since a force
feedback haptic interface can make up a large part of the cost
of a laparoscopy simulation station, it is important for surgical
educators to know how improvements in haptic rendering will relate
to improvements in training effectiveness.
Related
Research
Previous
work by other researchers includes research on task performance
of a simulated dissecting and suturing task at different magnitudes
of force feedback. The relationship between the quality of force
feedback and task performance has been researched in other tasks
but not in surgery. Since the use of our senses is very task specific,
these results do not necessarily apply to minimally invasive surgery.
Experimental
Question
How
does surgical performance change with each haptic interface quality
parameter (e.g. friction, cogging, maximum force etc.)?
Our
research is based on the understanding that there are three different
interpretations of haptic interface quality. The first interpretation
directly links the quality of the interface to the quality of
the components: if two haptic devices are exactly the same, but
one has motors with lower inertia, one would say the low inertia
device is of higher quality. A second interpretation of haptic
quality is that of subjective user tests: which of the two devices
'feels' better. This is different from the first interpretation
since, possibly dependent on the application, one might not be
able to tell the difference between the two previously mentioned
devices. A third interpretation is based on task performance.
Maybe with inexpensive components, the interface doesn't feel
as good, but the user is able to execute the task just as good.
For this project we're interested in task performance: surgical
performance in this case.
Surgical
performance measures
A
previously developed quantitative method [McBeth] that relates
instrument kinematics to surgical performance will be expanded
to include algorithms that take into account force and torque
values. The following aspects of surgical performance will be
included:
- Task completion time
- Tool kinematics
- Force and torque characteristics
- Error frequency
Hardware
degradation measures
Inexpensive
components make an inexpensive interface. We are modeling the
characteristics of inexpensive components such as motors, transmissions
and electronics, and incorporating them into the overall system
behavior. Using these models, we can make a high quality interface
emulate the characteristics of an inexpensive device.
For
flexibility, all hardware degradations are made in software.The
following picture shows where we intercept the signals to insert
our transfer functions.
References
Paul
B. McBeth, A.J. Hodgson, A.G. Nagy and K. Qayumi, “Quantitative
methodology of evaluating surgeon performance in laparoscopic
surgery”. MMVR, January 2002
K.E. MacLean, “Emulation of Haptic Feedback for Manual Interfaces”,
Ph.D. Thesis, MIT, 1996
Christopher R. Wagner, Nicholas Stylopoulos, Robert D. Howe, “The
Role of Force Feedback in Surgery: Analysis of Blunt Dissection.”
Symposium on Haptic Interfaces for Virtual Environment and Teleoperator
Systems 2002: 73-79
|
