Though the course of ongoing development of a dynamic "Virtual Rat", the need for physically based simulation has been established. To support such physical modeling, data regarding the material properties of various tissues is required. We present the results of in vitro testing of rat organ tissues and discuss preliminary comparisons of the results with a Finite Element simulation of the test scenario and a method to extract parameters from test data.
The mechanical properties of soft tissues are of increasing interest for diagnosis and surgical simulation. Techniques exist to acquire the stress-strain or strain rate behavior of biological tissue (e.g. elastography, force-displacement methods), but there are few experiments in which independent methods are directly compared.
The instruments are the TeMPeST normal indentation device (Simulation Group, CIMIT), and the ROSA-2 rotary shear device (University of Tuebingen). TeMPeST exerts small oscillatory forces through a 5mm right circular punch while recording applied load and relative displacement. ROSA-2 employs a 6mm circular contact which undergoes rotary oscillations relative to a concentric ring fixed to the tissue. Non-slip contact is maintained using a pin array, an abrasive surface or a tissue sealant/adhesive. A calibrated galvanometer exerts torque, while angular displacement is recorded. A preliminary experiment to verify non-slip static contact will also be described.
Static testing has verified non-slip contact using the needle array or the adhesive under both visual and fluoroscopic imaging (injected with contrast medium) on bovine liver in vitro. Histological analysis is being performed to study the limits of angular deformation without damage to the tissue. Both the TeMPeST and a previous version of ROSA have been used to measure tissue properties, and side-by-side in vivo testing will be performed shortly.
This paper describes initial work to determine whether indentation and rotary shear produce comparable results. Such results serve the dual purposes of verifying the measured visco-elastic behavior as well as the validity of the two independent measurement techniques.
The development of a Virtual Rat has been ongoing. During the development the need for physically based simulation has been established. We present the results of in vitro testing for the examination of soft-tissue properties, and compare the results with a Finite Element simulation of the test scenario. We also provide results for the use of these obtained material properties within a mass-spring simulation engine.
Surgical simulation is a field that shows much promise in terms of permitting surgeons to learn new skills, as well as enabling the development of new procedures and instruments while minimizing the use of animal subjects. Essential to simulation with realistic force feedback are material property data for the wide variety of soft tissues, both in healthy and diseased conditions. Properties are known to change significantly after death, so measurements should ideally be made in vivo. This paper reviews a number of methods in use to perform in vivo testing. It focuses on ongoing work using a minimally invasive instrument that uses normal indentation to acquire the force-displacement response of solid organ tissues such as liver, spleen and kidney. The design and characterization of the instrument as well as some of the in vivo and other test results will be presented.
Surgical simulation is a field that shows much promise in terms of permitting surgeons to learn new skills, as well as enabling the development of new procedures and instruments while minimizing the use of animal subjects. Essential to simulation with realistic force feedback are material property data for the wide variety of soft tissues, both in healthy and diseased conditions. Properties are known to change significantly after death, so measurements should ideally be made on living tissue. Further, tissue often has visco-elastic behaviors, so measuring properties over a range of frequencies is important. This paper presents the design of the TeMPeST 1-D (1-axis Tissue Material Property Sampling Tool) and its use in ongoing measurements of the visco-elastic properties of solid organ tissues.
Surgical technology and techniques have advanced at an increasingly rapid pace, beginning perhaps in the 1980s with the wider application of laparoscopic surgery, progressing up to the present day, where a growing number of hospitals in the United States are equipping themselves with telerobotic systems which permit minimally invasive heart surgery. With the new techniques and instruments comes a demand for additional training, both at the novices' and the experienced practitioners' levels.
To answer this demand, numerous surgical simulation systems, analogous to flight simulators for pilots, are being developed. These systems include such components as high resolution video feedback, tactile and haptic interfaces, and realtime simulation systems to calculate the deformations of tissue, as well as the force and visual feedback. Calculation of the deformations requires knowledge of mathematical models describing the behaviour of the tissues, together with the values of the parameters characteristic to each tissue type. Models and parameters may be different not only between different tissue types, but may also vary with age, with location on a given organ, with variations due to disease, and between individuals, amongst other influences.
This talk will focus on the wide variety of measurement methods that are in use and under development to permit acquisition of the data necessary to determine the parameters for tissues in vivo. Systems that will be reviewed include a number that are non-invasive, involving CT, MRI and ultrasound imaging, and surgically invasive methods that generally involve applying excitations directly to the tissue in question and examining the force-displacement response. Special attention will be given to work under way with the TeMPeST 1-D instrument, a minimally invasive device that is being used to acquire data from porcine liver and spleen in vivo. It also serves as a precursor to instrumentation that will be designed to acquire human tissue properties.
To support the ongoing development of software-based surgical simulation systems, work is underway to acquire the mechanical properties of living tissue. When such simulations include force feedback, visco-elastic properties must be evaluated over a range of frequencies relevant to human perception and motor control. A minimally invasive instrument has been developed which can perform normal indentation on solid organs, and apply and measure deformations over a frequency range from DC to approximately 100Hz. Measurement performance was validated on a series of objects and materials with known properties, and the device was subsequently used in in vivo tests on porcine liver. Results of these validation tests as well as the data extracted from the in vivo experiments are presented. Testing is ongoing, and will be expanded to more completely characterize liver, as well as porcine spleen and other solid organ tissues. While these animal tissue property tests are valuable in and of themselves, they pave the way for the development of instruments and experimental protocols suitable for the measurement of human tissue properties.
Surgical simulation systems need not only models that capture the behaviors of living tissue, but also parameters measured from real tissues to make such models meaningful. A portable system called the TeMPeST 1-D (Tissue Material Property Sampling Tool) that can acquire force-displacement responses in vivo has been developed. By fitting these data to the form of a chosen model, the tissue parameters can be obtained.
The data acquisition tool is suitable for minimally invasive or open surgical use. It measures normal indentation force-displacement response over a frequency range from DC to approximately 100Hz. This permits the investigation of the visco-elastic properties of living solid organ tissue. It can exert forces up to 300mN, and has a range of motion of ±500µm. The TeMPeST 1-D was used to measure the frequency-dependent stiffness of porcine liver in vivo in a proof-of-concept demonstration, and is being used in a more comprehensive series of tests. Based on simple tissue models, preliminary estimates for tissue stiffness are presented and the frequency-dependent and non-linear characteristics are discussed.
Surgical simulation, whether for training or for instrument or procedure prototyping, depends in part on knowledge of the material properties of the tissues in question. Generally speaking, tissue force-displacement responses exhibit non-linearity, anisotropy, inhomogeneity and time dependent features, some or all of which may be important to include in a tissue model. Further, tissue properties tend to change after death, so measurements made on living tissues are preferable to those made on excised samples. A number of devices have recently been developed to measure some of these properties for solid organ tissues in vivo, one of which is the TeMPeST 1-D.
The Tissue Material Property Sampling Tool 1-D is a minimally invasive surgical instrument which exerts small normal loads on tissues and simultaneously measures displacement and applied load. It is capable of generating vibratory stimuli up to approximately 80Hz, and has a range of motion of +/- 500um. Through careful selection of mean applied load, and frequency and location of stimulation, it can be used to investigate linear stiffness, non-linear effects, damping/viscous characteristics and inhomogeneity over the surface of an organ such as liver or spleen.
In my talk I'll introduce the TeMPesT 1-D system. I will discuss the geometric and other approximations, and modeling necessary for tissue parameter extraction. I'll go into some detail on the instrument hardware and describe experiments designed to verify the performance of the instrument on known elements and materials. The talk will conclude with discussion about in vivo testing on porcine liver and spleen, currently underway. I'll present the results of preliminary analysis of early experiments, and discuss some of the directions towards which our research is headed.
Medical simulation offers the opportunity to revolutionize the training of medical personnel, from paramedics and corpsmen to physicians, allowing early learning to occur in a no-risk environment, without putting patients at risk during the professional's early learning curve. However, the complexity of the problems involved in the development of the medical training systems as well as the spectrum of scientific fields that need to be covered have been a major limiting factor to the achievement of realistic simulations. We think that success in this effort cannot occur through uncoordinated efforts among domain experts working within their own fields. Success will come through medical personnel working side by side with engineers, computer scientists and designers to develop a simulation system that is useful and relevant. As part of our overall program to develop medical simulation, we have identified a critical infrastructure technology that will enable collaboration among simulation developers. When implemented, this Common Anatomy Modeling Language (CAML) will provide a common architecture for integrating the individual components of a medical simulation system.
Surgical training today, beyond what can be learned in didactic form or practice on animal or other models, is subject to the availability of appropriate training cases from which students can learn. This is especially true for battlefield surgery, as civilian hospitals may not expose doctors to frequent examples of relevant injuries. To provide a more uniform training experience, covering a standard suite of typical operations without relying on the misfortune of patients requiring surgery, many groups are developing computer-based surgical simulation systems.
One of the current areas of development is the implementation of force and tactile (haptic) feedback in simulations. To create a model with realistic haptic feedback, knowledge of the material properties of the tissues in question is essential. While there is much data from tissue samples in vitro, the properties of living tissue in situ are mostly unknown. From the data that is available, it is clear that living tissue and tissue in vitro can have radically different mechanical properties.
For this reason, our group is developing surgical tools that will be able to measure the force-displacement characteristics of a variety of tissues in living organisms. Taking these data over the range of frequencies relevant to haptic simulation provides information to extract stiffness and material damping parameters of different kinds of tissue.
The tools are being designed for use during minimally invasive surgery, but will permit data to be acquired either during MIS or open procedures. Animal tests are expected to commence in early 2000, but the tools are being designed with safety considerations in mind for eventual use in humans. Data will be taken both for solid organs and for selected elements of the vasculature. These data will be used in simulation systems under development at the Center for Innovative Minimally Invasive Therapy at Massachusetts General Hospital and the Laboratory for Human and Machine Haptics at MIT.
This paper describes the test-bed telesurgery system that was developed in the Human Machine Systems Laboratory, at MIT. This system was used to investigate the effects of communication time delays on controller stability and on the performance of surgical tasks.
The system includes a bilateral force reflecting teleoperator system, interchangeable surgical tools, audio and video communication between the master and slave sites, as well as methods to generate time delays between the sites. To compensate for the time delays, various control schemes were investigated, leading to the development and selection of Fuzzy Sliding Control (FSC). With a stable teleoperator system, experiments in performing a variety of surgical exercises were performed. These looked at the performance of a team of a telesurgeon and local assistant given a number of different time delay scenarios, including synchronous and asynchronous force and audio/video feedback.
Current efforts in surgical simulation very often focus on creating realistic graphical feedback, but neglect some or all tactile and force (haptic) feedback that a surgeon would normally receive. Simulations that do include haptic feedback do not typically use real tissue compliance properties, favoring estimates and user feedback to determine realism. When tissue compliance data are used, there are virtually no in vivo property measurements to draw upon.
Together with the Center for Innovative Minimally Invasive Therapy at the Massachusetts General Hospital, the Haptics Group is developing tools to introduce more comprehensive haptic feedback in laparoscopy simulators and to provide biological tissue material property data for our software simulation.
The platform for providing haptic feedback is a PHANToM Haptic Interface, produced by SensAble Technologies, Inc. Our devices supplement the PHANToM to provide for grasping and optionally, for the roll axis of the tool. Together with feedback from the PHANToM, which provides the pitch, yaw and thrust axes of a typical laparoscopy tool, we can recreate all of the haptic sensations experienced during laparoscopy. The devices integrate real laparoscopy toolhandles and a compliant torso model to complete the set of visual and tactile sensations.
Biological tissues are known to exhibit non-linear mechanical properties, and change their properties dramatically when removed from a living organism. To measure the properties in vivo, two devices are being developed. The first is a small displacement, 1-D indenter. It will measure the linear tissue compliance (stiffness and damping) over a wide range of frequencies. These data will be used as inputs to a finite element or other model. The second device will be able to deflect tissues in 3-D over a larger range, so that the non-linearities due to changes in the tissue geometry will be measured. This will allow us to validate the performance of the model on large tissue deformations. Both devices are designed to pass through standard 12mm laparoscopy trocars, and will be suitable for use during open or minimally invasive procedures. We plan to acquire data from pigs used by surgeons for training purposes, but conceivably, the tools could be refined for use on humans undergoing surgery.
Our work will provide the necessary data input for surgical simulations to accurately model the force interactions that a surgeon would have with tissue, and will provide the force output to create a truly realistic simulation of minimally invasive surgery.
A telesurgical model was developed for simulation experiments evaluating cooperative manipulation between a paramedic local to the patient and a physician operating through a telerobot. In this study we tested the hypothesis that sending the signals from a remote telesurgical setup asynchronously (sending the telemanipulator signals ahead of the video signals, which were delayed because of the time it took for compression/decompression) improve controller stability and favorably affect task performance by the medical team. We found essentially no difference in task performance between the synchronous and asynchronous transmission of the telesurgical signals when the physician operated the laparoscope and the assistant operated the laparotomy tools. But with asynchronous transmission we found a significant improvement (31% to 60%) in task completion time when the physician operated any of the laparotomy tools).
This paper reviews several interrelated experiments related to the improvement of systems by which to perform simple surgical procedures remotely using closed circuit video/audio and telerobotic manipulator devices over ISDN telephone communication channels. The realities of such technology include the existence of several second time delays, severe constraints on feedback bandwidth, and the lack of some desirable degrees of freedom for manipulation. Experiments were done to determine what sensory-motor tasks should be performed by the surgeon directly in a master-slave mode with haptic (position-force) feedback, what tasks should be programmed into and subsequently performed by a computer at the site of the patient, and what tasks should be performed by an untrained assistant (non-surgeon) physically located with the patient with the second-by-second supervisory guidance of the remote surgeon. Experiments were also done to find ways to ameliorate the instability in force feedback caused by the time delay. For each mode the paper identifies compromises in telepresence and sensory-motor performance tradeoffs between speed and accuracy. Based on experimental results, recommendations are made for ways to improve telesurgery systems now being developed.
A device that can present thermal stimuli to a virtual environment (VE) user has been developed for use with the PHANToM Haptic Interface. Using a thermo-electric heat pump and a water cooled heat sink, it can achieve heating and cooling rates of +11°C/s and -4.5°C/s, respectively, and can display temperatures over the entire comfort range of human sensation (5 to 45°C). The ThermoStylus can display the thermal characteristics of a VE, or represent any other scalar data by substitution (e.g. pressure, concentration, etc.). Applications include improved medical simulation, increased VE realism and improved physical simulations (e.g. thermo-fluid systems with flow and thermal variations). This stylus-style interface is both familiar to PHANToM users, and intrinsically safe. Water cooling prevents build-up of heat in the heat sink, and should a failure occur, the user can easily release the stylus. Current performance and future developments, including the determination of real thermal transients, are discussed.
This paper respresents the first conference at which my Ph.D. research was presented. You can find some more information at my haptics research page
One aspect of applying teleoperators to a surgical scenario that has been overlooked to date is how best to coordinate the actions of a tele-surgeon and an assistant on the scene of the operation. This research uses laparoscopic surgery as a model, and examines performance of the surgeon/assistant pair on simulated surgical tasks under different conditions of time delay and surgical tool assignment. Our experiments suggest that under time delay conditions, tasks are completed most quickly when the tele-surgeon controls only a laparoscopic camera, and instructs the assistant in how to complete the tasks. This paper also describes the teleoperator system used and a proof-of-concept demonstration that will be conducted between MIT and Massachusetts General Hospital in late 1996. The demonstration will make use of three ISDN telephone lines to transmit audio, video, and control signals.
The severe illness of one of our group members and departure of others has put the MIT/MGH demo on hold...more info when it becomes available
Many groups have used the PHANToM Haptic Interface arms to interact with virtual environments, but they can also be used as both master and slave in a teleoperator system. In the Human-Machine Systems Laboratory at MIT we are involved in a project to examine a number of issues related to the development of a telesurgery system. We have used a pair of "tool handle" PHANToMs to provide force feedback to the telesurgeon and as a platform for a dedicated telesurgical tool. The telesurgery system also provides for audio communications between the surgeon and an assistant at the remote site, and transmission of a color video image of the surgical task. The issues that we investigated were control of the teleoperator when subject to system time delays and the interaction between the telesurgeon and assistant under various time delay conditions. To support these efforts, the system includes hardware and software to generate delays, and surgical simulators to test the performance of the telesurgeon/assistant team.
The system and the experiments will be described, and the results and our experience will be discussed.
Teleoperator systems are beginning to be used in surgical settings, from telerobotically controlled laparoscopic cameras in clinical use to experimental systems which give a surgeon remotely controlled surgical tools. What has not been formally investigated is cooperation between a telesurgeon and a less skilled, but local, assistant. A telesurgeon will be subject to the limits of the teleoperator--imperfect sensory feedback (audio-visual, haptic) and time delays between the master and slave, for example. The assistant, unimpeded by these limits, does not have the knowledge and skills to perform an operation alone. By cooperating, tasks that could not be completed separately are possible.
This presentation describes our experiments to determine how best to coordinate the efforts of the surgeon and assistant under different communication time delays. We developed a one-armed, force reflecting teleoperator system equipped with a teleoperated laparoscopic surgical tool. A surgeon used it to perform a set of surgical training tasks together with a less skilled assistant. Round-trip time delays of 0, 0.6 and 1.2 seconds were generated in the haptic, audio and video feedback to the surgeon. The best way to assign up to three tools between the partners was determined based on task completion time. Lowest times occured when the surgeon controlled only the laparoscopic camera and a video pointer device, directing the assistant in the use of any other tools. This was especially true for non-zero delays, when completion time could be as much as six times longer for the surgeon using time-delayed forceps or shears instead of the laparoscope. However, even under 1.2s delay, completion times remained nearly constant when the assistant controlled forceps, shears or clip appliers under the surgeon's direction. These results suggest that until very high bandwidth communication channels are readily available, the best way to bring a remote surgeon's skills to a casualty, whether on a battlefield or in locations far from medical centers, may be by directing the hands of an assistant already on the scene.
This abstract describes a poster presentation I gave at the 40th HFES annual meeting.
The paradigm of telesurgery is characterized, where the remote surgeon not only manipulates surgical tools via a telerobot, but also provides supervisory control through a constrained communication channel to a human assistant who in turn manipulates the patient. In this sense the human assistant becomes a telerobot.
A work combining some of the early results from the telesurgery group with work by Suyeong Kim on human intervention in supervisory control
The HMSL is involved in a nation-wide project to develop a system which will permit remote monitoring and treatment of sick or injured persons. The aspect of the project upon which this paper will focus is the concept of "cooperative telesurgery"--the interaction between a surgeon acting through a teleoperator system and an assistant local to the patient. Cooperation is seen as a way to improve surgical performance because the surgeon has the skill and knowledge that the assistant lacks, while the assistant is not encumbered by the shortcomings of the teleoperator (i.e. time delay, limited visual/audio/haptic feedback, limited reach/d.o.f.'s). To date, there is very little published research on cooperation in task performance and virtually none on skilled manual tasks. This paper will review some of the available literature, discuss the cooperative telesurgery project in more detail and present the work completed and experiments planned for the future.
This was the first report describing my Master's degree thesis work. My Master's thesis advisor was Dr. T.B. Sheridan.
The medical field, and surgeons in particular, are turning to engineers to develop systems that help them learn their craft better. Mannequin-based systems, animal labs and surgery on cadavers each have drawbacks that could be addressed through realistic computer-based surgical simulation systems. To generate a simulation that includes both tactile/haptic and visual feedback, one must know what the material properties of tissue are, so that a finite element or other model can generate the proper predictions for interactions between surgical instruments and tissue.
This thesis presents the design, construction, characterization, and use of a minimally invasive surgical instrument designed to measure the linear visco-elastic properties of solid organs. The Tissue Material Property Sampling Tool, or TeMPeST 1-D, applies a small amplitude vibration normal to the surface of an organ such as liver or spleen, and records the applied force and displacement. It has a range of motion of up to 1mm, and can apply up to 300mN force with a 5mm right circular indenter. The open loop bandwidth of the system is approximately 100Hz, which is greater than the bandwidth of both the human visual and motor control systems.
The relationships between indentation force and displacement and material properties such as the elastic modulus of tissue are presented, and models are developed that show the expected response to a standard tissue model. Characterization and calibration tests demonstrate the response of the prototype components. Experiments performed on spring and mass elements and on silicone gel samples, which mimic tissue response, show that the TeMPeST 1-D can accurately measure their force-displacement responses.
The TeMPeST 1-D and its data acquisition system are intended to be portable, to be easily transported to and used in an operating room. The system was used in proof-of-concept experiments performed on live pigs; an example of the measured properties of porcine liver is presented.
The TeMPeST 1-D is the first in a series of instruments that will be developed to support the generation of a comprehensive atlas of tissue material properties.
My doctoral thesis advisor was
Prof.
J. Kenneth Salisbury.
This thesis presents the development and results of experiments studying the interaction between a telesurgeon and an assistant local to a patient. In particular, a number of simulated laparoscopic surgical tasks were performed under various conditions of teleoperator time delay and tool assignment between the surgeon and assistant. The time delays ranged from no delay to 1.2 seconds. Tools used in the tests included a laparoscope, laparoscopic grippers, shears and clip appliers.
To provide a setting in which to perform the experiments, a telesurgery system was assembled. The central component is a three degree of freedom, bilateral force reflecting teleoperator with a two d.o.f. teleoperated surgical tool without force feedback. The surgical tool can actuate a laparoscope, laparoscopic grippers or shears. Other components include two surgical patient/task simulators, one-way video and two-way audio communications hardware and a delay generation device.
The experimental results obtained using this system suggest that under delay conditions, telesurgeons should act only by controlling the laparoscope and instructing, while delegating performance of the manual tasks to the assistant. Improvements to surgical teleoperators could allow surgeons to interact directly for low time delays, but, depending on the cost of increased completion time, delegation may always be necessary for higher delays to obtain the best performance.
Some of the details of this research are described on the HMSL
Telesurgery pages
Click here
for MIT Libraries digitized version
Introduction
The objective of this manual is to provide the reader with technical
information about the work done to bypass VAL II (Victor's Assembly Language).
This report describes in detail the modification work that was done on
the robot controllers and is not intended as a troubleshooting guide.
The two industrial robots available in the Flexible Manufacturing Centre are the PUMA 560 robot and the AdeptOne robot. The controller hardware for each robot uses VAL II as the language/operating system for moving the arm, storing and executing programs, and numerous other tasks. The robots in the FMC lab provide an excellent experimental setup for advanced control schemes provided that the limitations of VAL II are removed. Therefore, a new controller needs to be constructed so that researchers have access to the position and motor control signals of the two robots.
To increase the flexibility of the controllers at the FMC lab, a bypass of VAL II and the servo hardware was necessary. One restriction imposed on the modification work was that the original VAL II controllers had to be left intact so that students and researchers could use either the original controller (with VAL II) or the new modified controller (programmed in C or Fortran) with the robots.
I helped to design, and built: a motor torque feedback board and a board to provide software control of arm power for the PUMA; and a power amp protection circuit and a gripper control circuit for the Adept. This report also describes the power-up calibration hardware.
A method of power-up calibration has been developed for the AdeptOne robot in the Flexible Manufacturing Systems laboratory. This will permit the determination of the position of each of the links of the robot, an ability which is lost when the robot is connected to a new VME-based controller. The technique involved the installation of optical sensors and the construction of a simple analog to digital converter to interface the sensors with the robot controller. The hardware and the algorithm which was developed to execute calibration are described.
This report describes the work done to complete the requirements of the undergraduate thesis course in mechanical engineering at McMaster University. My supervisor was Dr. H.A. ElMaraghy.
Executive Summary
Transmission cable galloping is a type of flow induced vibration which
can occur when ice builds up on cables that are exposed to certain wind
conditions. The cost to electrical utilities and consumers ranges in the
hundreds of thousands of dollars annually. I studied the aerodynamic properties
of models of cables with and without ice. I also examined the results of
some other researchers to check their validity. The results of my measurements
can be used to develop better computer models of how galloping occurs,
with the long term goal of reducing the damage and the associated costs.
This report satisfied a requirement for the final year Engineering and Management course at McMaster. It covers work done through the NSERC Undergraduate Student Research Assistantship award program, under the direction of Dr. Ö.F. Turan.