Development of an Anatomy Identification skill set for Laparoscopic and Robotic Minimally Invasive Procedures.
By
Ankur R Baheti
Department of Mechanical and Aerospace Engineering State University of New York at Buffalo Buffalo, New York 14260
A thesis submitted to the faculty of the Graduate School of The State University of New York at Buffalo in partial fulfillment of the requirements for the degree of Master of Science September 2008 1
2
Contents Abstract……………………………………………………………………………...................6
List of figures…………………………………………………………………………………...7 List of tables………………………………………………………………………....................9 1. Introduction…………………………………………………………………………………10 1. Minimally Invasive Surgery………………………………………………………...10 2. Drawbacks of Minimally Invasive Surgery………………………………………....11 3. Robotic Surgical System……………………………………………………………12 a. Zeus Robotic System……………………………………….......................13 b. The daVinci Surgical System (DVSS)……………………………………..14 i. The Master Console (Surgeon Console)…………............................15 ii. Patient side cart……………………………………….....................16 iii. EndoWrist Instruments……………………………........................16 iv. Vision system………………………………………………….…..17 4. Advantages of Robotic Surgical Systems…………………………………………17 5. Need for a Simulator for the DVSS……………………………………………… 18 6. Objective of the Thesis…………………………………………………………….19 2. Surgical Simulators………………………………………………………………………...20 1. History………….………………………………………………………………….20 2. Virtual reality based Laparoscopic Simulators…………………………………….21 a. LapSIM by Surgical-Science………….……………………………………21 b. Minimally Invasive Surgical Trainer (MIST-VR) by Mentice Inc.……….25 c. LapVR Virtual Reality System by Immersion Medical……………………27 3. How effective is VR for Surgery?............................................................................30 3
a. Validation Studies on VR Simulators…………………………………….30 i. MIST-VR……………………........................................................31 ii. LapSIM…………………………………………………………...32 4. Development of Cognitive Skills for Laparoscopic Training.................................33 5. Simulators for the daVinci Surgical Simulators (DVSS)…...……………………34 a. SurgicalSim Education Platform (SEP)………………………………......34 b. daVinci Surgical Simulator by the Chinese University of Hong Kong…..35 c. Institute for High Dimensional Imaging, The Jikei University School of Medicine, Tokyo, Japan…………………………………………………..37 d. The dV Trainer by Mimic Technologies………………………………….39 6. Redesign of the robotic surgical simulator (RoSS) for the DVSS………………..41 3. Development of Hardware Interface……………………………………………………..44 1. Design of the Frame…............................................................................................45 2. Design of the Wrist…………………………………………………………..........47 4. Development of Graphical Interface…………………………………………….............49 1. Mapping using Inverse Kinematics…………………………………………...…49 2. Omni Interface….……………………………………………………………….52 3. 3D Graphics rendering…………………………………………………………..52 4. Features…………………………….....................................................................52 5. Development of Skill Sets………………………………………………………54 a. Development of Psychomotor Skill Sets………………………………...54 i. Pick and Place…………………………………………………....54 ii. Traversal………………………………………………………....55 iii. Stretching and Clipping…………………………………………55 b. Development of Cognitive Skill Sets…………………………………..56 i. Cauterization……………………………………………………..57 ii. Bleeding Control…………………………………………….. ....57
4
5. Experimental Details………………………………………………………………………60 1. Experimental Protocol……………………………………………………………..60 2. Experimental Results………………………………………………………………61 3. Results of the Post-Task Survey…………………………………………………...65
6. Conclusions and Future Work…………………………………………………………….67 Appendix…………………………………………………………………………………….69 References…………………………………………………………………………………...73
5
Abstract This thesis involves the development of a robotic surgical simulator (RoSS) for the daVinci surgical system (DVSS) increasing the size of the psychomotor skill set and adding a module for the improvement of cognitive skills. This thesis deals with the redesign of the existing RoSS system with a more compact frame and a newly designed wrist that preserves the kinesthetics of working on the actual system. The cognitive skill sets developed are basic skill builders of anatomy identification and bleeding control.
6
List of Figures Figure 1: Setup of a typical laparoscopic surgery……………………………………………..11 Figure 2: Zeus robotic system…………………………………………………………………13 Figure 3: daVinci Surgical System (DVSS)…………………………………………………..14 Figure 4: Surgeon‟s Console…………………………………………………………………..15 Figure 5a: Master Controls (left) and the Foot Pedals (right)…………………………………15 Figure 6: Patient Cart Showing two instrument arms and an endoscope arm…………………16 Figure 7: EndoWrist Instruments……………………………………………………………...17 Figure 8: Camera Navigation Task in LapSim………………………………………………...21 Figure 9: Instrument navigation task in LapSim………………………………………………22 Figure 10: Coordination Task in LapSim……………………………………………………...22 Figure 11: Grasping Task in LapSim………………………………………………………….22 Figure 12: Cutting Task in LapSim……………………………………………………………23 Figure 13: Clip applying task in LapSim………………………………………………………23 Figure 14: Lifting and grasping task in LapSim……………………………………………….24 Figure 15: Precision and Speed Task in LapSim………………………………………………24 Figure 16: Setup of MIST-VR…………………………………………………………………25 Figure 17: Tasks in modules Core Skills 1 and Core Skills 2 of MIST-VR…………………..27 Figure 18: Camera Navigation task of Lap VR………………………………………………..28 Figure 19: Peg Transfer Task in Lap VR……………………………………………………...28 Figure 20: Cutting Task in Lap VR……………………………………………………………29 Figure 21: Clip applying task in Lap VR………………………………………………………29 Figure 22: Setup of Surgical Education Platform……………………………………………..34 Figure 23: Setup of the daVinci simulator of Chinese University of Hong Kong…………….35 Figure 24: Electromechanical Gripper…………………………………………………………36 Figure 25: Bean drop drill in daVinci Simulator of Chinese University of Hong Kong………37 Figure 26: A patient organ model on the left and its corresponding sphere filled model on the right……………………………………………………………………….…38 7
Figure 27: A Tele-surgical simulator for the DVSS by Jikei University school of Medicine…………………………………………………………………………39 Figure 28: The dV Trainer developed by Mimic Technologies……………………………...40 Figure 29: Ball Drop…………………………………………………………………………41 Figure 30: Cylinder capping………………………………………………………………….42 Figure 31: Setup of DVSS…………………………………………………………………...44 Figure 32: daVinci master (left) and Phantom Omni mounted upside down (right)…………45 Figure 33: The Omni mounted on the frame…………………………………………………45 Figure 34: Setup of the frame structure………………………………………………………46 Figure 35: The Simulator……………………………………………………………………..47 Figure 36: The wrist………………………………………………………………………….47 Figure 37: The line diagram of the master controller of DVSS (left) The line diagram of the master controller of RoSS (right)………………………..49 Figure 38: The complete RoSS system………………………………………………………53 Figure 39: Pick and Place…………………………………………………………………….54 Figure 40: Traversal…………………………………………………………………………..55 Figure 41: Stretching and Clipping…………………………………………………………..56 Figure 42: Anatomy Identification…………………………………………………………..57 Figure 43: Bleeding Control ………………………………………………………………...58 Figure 44: The box plots for the mean times for completion of the test…………………….62 Figure 45: The box plots for the mean number of errors for the test………………………..63 Figure 46: The box plots for the mean number of correct answers for the test……………..64 Figure 47: The rating of the anatomy module……………………………………………….65 Figure 48: The rating of the bleeding module……………………………………………….66 Figure 49: The overall rating of the RoSS system…………………………………………...66
8
List of Tables Table 1: DH parameters of DVSS master………………………………………………….50 Table 2: DH parameters of RoSS master…………………………………………………..51 Table 3: Means of the various parameters…………………………………………………60
9
Chapter 1: INTRODUCTION Robotic technology made significant progress in the field of medicine in the 1990‟s. Since then there has been an increase in the application of robots for surgical procedures. They have provided the surgeons with higher degree of precision while performing the minimally invasive surgeries. Minimally invasive surgery basically involves the making of small incisions in the body, and using instruments and a camera to perform the surgery. Doctors have found that Robotics based minimally invasive procedures have certain advantage over traditional Laparoscopic surgery. Because of this robotics is growing fast as an alternate to traditional procedures. Robotic surgeries have been around for almost a decade. Almost all the urologic procedures are performed by the traditional minimally invasive technique manually can also be performed robotically.
1. Minimally Invasive Surgery Surgery has been a part of medical practice for centuries. In conventional open surgeries, large incisions are made on the body of the patient during surgery. This results in increased recovery time, pain, and discomfiture for the patient. Also, the risk of heavy bleeding and other complications are high in conventional open surgeries. This called for new techniques to reduce or to completely eliminate these problems. This led to the introduction of a new technique of surgery called the Minimally Invasive Surgery (MIS) in the late 1980s.
Minimally invasive surgery is a revolutionary technique that has been applied in clinical practice for a variety of surgical procedures. In this method, surgeon performs the operation with the help of a small endoscopic camera and several long, thin rigid instruments through natural body openings or small surgical incisions, as compared to large incisions in conventional open surgery. This is otherwise called as keyhole surgery or laparoscopic surgery. Figure 1 shows the setup of a typical laparoscopic surgery.
10
Figure 1: Setup of a typical laparoscopic surgery ([1])
The advantages of MIS in comparison to the open surgery are; less postoperative pain, faster recovery, minimal blood loss and better postoperative immune function.
2. Drawbacks of Minimally Invasive Surgery Several limitations exist for MIS; most of them are due to technical and mechanical nature of the equipment. The instruments don‟t transfer the forces to the hand. Hence, the surgeons operating cannot gauge the force that is being applied. This makes tissue manipulation dependent only on visualization. The surgeon does not get any force feedback apart from that transferred by the rigid tool.
Laparoscopic surgery is a minimally invasive procedure. The instruments such as the camera, the tools etc are inserted into the body through holes of very small diameter. Due to this the instruments have only 4 degrees of freedom; two rotations about the point of insertion, one rotation about the axis of the tool and one translation along the length of the tool. The surgeon‟s mobility is greatly reduced as the human hand has higher number of degrees of freedom when compared to the tool.
11
Moving the instruments while watching a 2-dimensional video monitor is counterintuitive. Furthermore, the surgeon must move the instrument in the opposite direction from the desired target on the monitor to reach the site of interest. This is called fulcrum effect [2]. Therefore, the natural hand-eye coordination is lost. Finally, tremors are readily transmitted from the surgeon‟s hands through the length of the rigid instruments, making it extremely difficult to do any precise surgical task. [3].
The motivation for the development of robotic surgery is to overcome the limitations of the current laparoscopic technologies and to reap the benefits of minimally invasive surgery.
3. Robotic Surgical Systems The first ever use of a robot in a laparoscopic surgery was the use of the electromechanical endoscope guidance system. The first robotic arm applied for this purpose was launched in 1994 called AESOP (Automated Endoscopic System for Optimal Positioning). The camera control in this system was in the hands of the surgeons [4-5]. This system could be controlled and manipulated using a foot pedal or voice control. “Since its introduction more than 10 competing systems have come to the market. The basic principles vary from completely mechanical to electromagnetic, and control is achieved by using joysticks, foot pedal, voice control, head-mounted light emitting signal systems, or even eye ball tracking” [6-8]. These robotic camera guidance systems are extremely useful while doing lengthy procedures and under circumstances with limited human assistance. They enable easy access to areas which are inaccessible.
After the successful introduction of robotic assisted camera guidance system, research programs were started to improve the dexterity of the surgeon in minimally invasive surgery. The concept of telemanipulation surgery was being tried out to perform surgeries on soldiers from a remote safe area near the battlefield. This concept didn‟t materialize because of transmission problems 12
(loss of data or lag) and ergonomical issues. However, this project resulted in the development of prototype systems that could be very valuable for working over a short distance with improved dexterity and vision for the surgeon [9].
Two commercial systems for performing robotic telemanipulation surgery came in to the market in 1998: The Zeus system (Computer Motion, Goleta, CA) and the daVinci system (Intuitive Surgical, Mountain View, CA). Both the companies initially aimed at supporting minimally invasive cardiac surgery, but have expanded their field of interest to laparoscopic procedures in gastrointestinal, gynecologic and urologic surgery.
a. Zeus Robotic system Zeus robotic system was an improvement made by Computer Motion Inc. to the already existing product AESOP robot [10]. Aesop robot was able to support and position endoscopic cameras. With Zeus, all the instruments are robotic. The surgeon as seen in figure 2 can sit comfortably at a master console and control the slave robotic instruments using a pair of master manipulators looking at a 2-D screen.
Figure 2: Zeus robotic system. 13
The slave has two instrument arms and one camera arm. The camera of the system is controlled by voice. The instrument has 5 degrees of freedom, as compared to 4 degrees of freedom in the conventional laparoscopic surgery. However, this product has been discontinued and is not available in the market.
b. The daVinci Surgical System(DVSS) The daVinci surgical system (DVSS) was introduced by Intuitive Surgical Inc. in 1999, and was superior to Zeus in many ways, offering superior vision system, increased dexterity by its EndoWrist instruments, an extra electromechanical arm to hold and manipulate instruments. Figure 3 shows the DVSS.
Figure 3: daVinci Surgical System.
The Components of the daVinci Surgical System include,
14
i. The Master Console (Surgeon Console) Surgeon performs the surgery comfortably seated in the console, thereby reducing surgeon‟s fatigue. Figure 4 shows the view of surgeon‟s console.
Figure 4: Surgeon’s Console The surgeon‟s fingers grasp the master controls shown in figure 5a. The system translates the hand movements into precise real-time movements of the surgical instruments inside the patient. Hand (holding the master control) and eye (seeing through the stereoscopic lens) are aligned, providing better ergonomics than conventional laparoscopy.
Figure 5: Master Controls (left) and the Foot Pedals (Right)
The foot pedals form a very important part of controlling the system. The foot pedals consist of the clutch, camera, electrocautery etc controls. Clutch helps to bring the hands of the surgeon in 15
a comfortable position while operating. Camera controls the position of the camera inside the patient. Electrocautery provides the current needed during the cauterization process. The arrangement of foot pedals in the DVSS is as shown in figure 5b.
ii. Patient Side cart The patient side cart has four electromechanical arms. Three of these are used for holding the tools and the fourth one is used for mounting the camera. The third arm of the patient side cart is used only during complicated procedures like suturing. Also, it can be used for providing counter traction. This would eliminate the need for a cart-side surgeon. The fourth arm holds the camera whose position is controlled by the surgeon using the foot pedals and the robotic arms of the master console. Figure 6 shows two robotic arms and one endoscope arm.
Figure 6: Patient Cart Showing two instrument arms and an endoscope arm
iii. EndoWrist Instruments (Laparoscopic Tools) The EndoWrist instruments consist of a set of about ten instruments. Each of these instruments is used for a specific task during the surgery such as grasping, suturing, cauterizing, clipping etc. The instruments are as shown in figure 7. All of the instruments are designed the same way and each of them has seven degrees of freedom.
16
Figure 7: EndoWrist instruments [11]
iv. Display The display system consists of a dual lens endoscope and image processing equipment that provides a high resolution 3D image of the operative field. Image processing equipment comprises of specialized edge enhancement and noise reduction equipment. There are three different endoscope cameras that provide for different view of the same scene. The different endoscope cameras are the 0 o lens, 30o inwards lens and the 30o outwards lens. Generally a 0 o lens is used during the surgery. The other lenses are used for the tool and trocars insertions.
4. Advantages of Robotic Surgical Systems
Improves dexterity of the surgeon due to increased degrees of freedom of the instruments used.
These systems scale down large movements of the hand in to micro movement of the instrument, thereby improving precision.
Hand-eye coordination is intuitive, thus eliminating fulcrum effect.
Restores ergonomic position for the surgeon.
Tremors of the surgeon‟s hands are filtered out and do not reach the end-effecter.
3D view along with depth perception is much better than the view obtained using standard laparoscopic cameras. 17
5. Need for a Simulator for the DVSS The DVSS is most widely used system among the available surgical robots. It is commonly used for the surgical procedures like prostatectomy (removal of prostate), cardiac valve replacement and cholecystectomy (removal of gall bladder). Due to numerous advantages of the DVSS, the number and type of surgeries performed using this is increasing rapidly. As more institutions are buying this robot, it becomes necessary to develop a trainer which is less expensive and yet trains effectively preserving the kinesthetics and also having all the functionalities possessed by DVSS.
The DVSS (Intuitive Inc. Sunnyvale, CA) is the only commercially available robotically-assisted surgical system in the market at this time. More than 500 DVSS‟s are in place in hospitals and other institutions around the world [11]. With the use of high precision devices like these, the surgeons can perform many surgical tasks with more precision and repeatability [12]. However, robotic surgery requires extensive training and generally the learning curve is very steep, due to the high degree of dexterity and visualization required to carry out the techniques. [13].
A key challenge for surgical training is to provide conditions for effective learning without putting the patient‟s health at risk. Robotic surgery simulators makes it possible by providing safe and realistic learning environment and reducing the time and cost of training [14]. Simulators also cater to different levels of difficulty. They not only train young surgeons acquire new skills, but also help experts maintain their proficiency during absences from the operating room.
Due to the increased use of surgical robot (daVinci robot; Intuitive Surgical, Sunnyvale, CA) to perform minimally invasive surgical procedures, there is a need for a valid VR-assisted robotic surgery simulator. This would help reduce the steep learning curve associated with many of the complex procedures and thus enable better outcomes. To date, such simulators do not exist [15]; however, several agencies and corporations are involved in making this a reality.
18
6. Objective of the Thesis The objective of this thesis is to improve the simulator designed for the daVinci Surgical System in the first phase of this project. The first system developed skill sets for improving the motor skills of the surgeon. This work involves the enhancement of already developed skill sets and also the development of a cognitive skill sets which could be used to train both the novices and the experts.
This project was taken up in collaboration with Rowell Park Cancer Institute where DVSS is being used for doing a wide variety of laparoscopic surgical procedures. It is joint initiative of Virtual Reality Lab, University at Buffalo and Roswell Park Cancer Institute.
19
Chapter 2: Surgical simulators 1. History Simulators can be classified into three types: those that use physical models, computer -based virtual reality simulators and hybrid simulators that combine physical and virtual methods.
The most primitive physical model based simulators depended on cadavers and live animals [16], but their use was limited due to ethical reasons and high costs. With the advancement of medical technology, inanimate models were being used in training simulators. For instance, a user can practice insertion on a synthetic skin model [17]. These simulators require extensive support from experts. Hybrid simulators have been developed to overcome this disadvantage, but they are still in early stages of development [18]. With the development in the field of computer graphics and virtual reality in the 1990‟s computer based simulators started gaining importance. The first virtual reality simulator was developed for leg surgery [19]. This system enabled the user to practice surgical manipulation of a virtual leg. The application of virtual reality to the laparoscopic surgery started with the development of a surgical simulator for laparoscopic cholecystectomy [20]. Later on, many commercial products were developed for training surgeons in laparoscopic surgery skills. These simulators trained the users to acquire basic laparoscopic skills like navigation, touching, grasping, stretching and translocation. Some of these also trained the users on advanced skills such as dissection, diathermy, suturing and knot-tying. The most widely used laparoscopic skills trainer is the Minimally Invasive Surgical Trainer (MIST-VR) developed by Mentice Inc [21]. Other prominent ones for laparoscopic surgery are the Lap VR simulator [22] developed by Immersion Medical Inc. and the LapSim simulator [23] developed by Surgical Science Inc.
20
2. Virtual Reality based Laparoscopic Simulators Before designing a virtual reality simulator for the DVSS, it is imperative to study the various virtual reality based laparoscopic surgical simulators thoroughly for various facets like, the hardware design, the computer interface, the basic skill sets, advanced skill sets and any other special features that have been added to increase the functionality. The following section studies the most widely used virtual reality based laparoscopic surgical simulators.
a. LapSIM by Surgical-Science LapSIM [23] is one of the most widely used laparoscopic simulators. The module of LapSIM consists of the following tasks,
1. Camera Navigation This task helps the user learn the skill of camera navigation. Figure 8 shows the camera navigation task where the user has to move the yellow circle around the ball.
Figure 8: Camera Navigation Task in LapSim.
2. Instrument Navigation This task trains the user in moving two laparoscopic instruments in 3 dimensional space guided by a 2-D image in a precise manner as shown in figure 9.
21
Figure 9: Instrument navigation task in LapSim.
3. Coordination This task trains the user to use both the hands in a coordinated fashion, one for holding the camera and other for navigating the instrument as shown in figure 10.
Figure 10: Coordination task in LapSim.
4. Grasping After the user learns the basic navigation and coordination skills, the user learns the task of grasping an object. Figure 11 shows the task of grasping the red ball.
Figure 11: Grasping task in LapSim. 22
5. Cutting
This task trains the user to cut with a forceps or ultrasonic scissor as shown in figure 12.
Figure 12: Cutting task in LapSim.
6. Clip applying This task involves the use of both the instruments simultaneously, one for holding the vessel and other for clipping as shown in figure 13. Available instruments are clip applier, grasper, scissors and a suction device.
Figure 13: Clip applying task in LapSim.
7. Lifting and grasping
Here, the grasping exercise is taken one step further. An object with tissue-like properties has to be lifted in order to remove an object as shown in figure 14. 23
Figure 14: Lifting and grasping task in LapSim.
8. Precision and Speed This task also trains the user on navigation by using a game like approach. In the task the user has to hold the balls marked L with the left hand tool, and those marked R with the right hand tool, one after another. The task is as shown in figure 15.
Figure 15: Precision and speed task in LapSim.
Besides these basic tasks, the simulator also trains in suturing and fine dissection. In addition to this, LapSim has a procedural trainer for Cholecystectomy and various gynecological procedures like tubal occlusion, salpingectomy, tubotomy and myoma suturing [23].
b. Minimally Invasive Surgical Trainer (MIST–VR), by Mentice Inc. MIST–VR is the most studied virtual reality simulator. The system comprises a frame holding two standard laparoscopic instruments electronically linked to a computer. The screen displays 24
the movement to the surgical instruments in real time 3D graphics. The setup of the MIST VR simulator is as shown in the figure 16.
Figure 16: Setup of the MIST-VR
The simulator has 2 modules Core Skills 1 and Core Skills 2. Core Skills 1 and 2, each has 6 Laparoscopic Skill tasks.
Core Skills 1
1) Task 1 - Acquire Place: This task involves acquiring a sphere using one hand and placing it in a specified 3D location. 2) Task 2 - Transfer Place: This task involves acquiring a sphere using one hand, transferring it to the other hand and placing it in a specified 3D location. 3) Task 3 - Traversal: This task involves traversing the length of the target object (cylinder) using hand-over-hand transfer. 4) Task 4 - Withdraw-Insert: This task involves withdrawing a tool from the operating space and reinserting it accurately. 5) Task 5 - Diathermy: This task involves applying diathermy to a specified target object at a specified 3D location 25
6) Task 6 - ManipDiathermy: This is the most difficult task in Core Skills 1 module. The task involves acquiring an object (sphere) and applying diathermy to a target on the surface of the object, maintaining the object in a specified 3D location.
Core Skills 2 1) Task 1 – SC Stretch: This task involves stretching a cylindrical object to a predetermined length using one hand. 2) Task 2 – SC Clip: This task involves clipping a specified target location on a cylindrical object using one hand. 3) Task 3 – SC Stretch Clip: This task involves stretching a cylindrical object to a predetermined length and clipping the center section of the same. 4) Task 4 – SD Stretch: This task involves stretching a cylindrical stretch diathermy object using one hand to a predetermined length. 5) Task 5 – SD Diathermy: This task involves applying diathermy along the surface of a cylindrical object. 6) Task 6 – SD Stretch Diathermy: This task involves stretching a cylindrical object to a predetermined length using one hand, and applying diathermy along a predetermined line along the object.
Each task is based on a key surgical procedure employed in laparoscopic cholecystectomy, as shown in figure 17, using simple geometrical shapes rather than tissue to allow the trainee to concentrate on the development of key psychomotor skills.
26
Figure 17: Tasks in modules Core Skills 1 and Core Skills 2 of MIST-VR [21]
Performance is measured by time, number of errors, and the efficiency with which the exercise is performed. The system provides data on the performance of groups and individual trainees, thereby enabling comparative analyses between groups and objective assessment of individuals [21].
c. LapVR™ Virtual Reality System, by Immersion Medical Lap VR simulator has an essential skills module that consists of four skill tasks, Camera Navigation, Peg Transfer, Cutting and Clip Applying. Each of these tasks are described below.
1. Camera Navigation This task involves moving the capture ring over a series of target of numbers. This task helps the user to navigate a camera and maintain focus on the target object in a dynamic environment. The camera navigation task is as shown in figure 18. 27
Figure 18: Camera Navigation task of Lap VR
2. Peg Transfer The task involves picking a series of pegs from the ground and inserting it inside a hole on a pegboard. This task helps the user to improve instrument navigation, hand31 eye coordination and grasping. It also sharpens the depth perception of the user. The peg transfer task is as shown in figure 19.
Figure 19: Peg Transfer Task in Lap VR
3. Cutting The task involves cutting the cloth along a pre-marked boundary, making use of a scissor for a cutting, and grasper for holding the cloth to create traction as shown in figure 20.
28
Figure 20: Cutting Task in Lap VR
4. Clip applying The task involves applying clips at four pre-marked locations on the artery as shown in the figure 21.
Figure 21: Clip applying task in Lap VR
Apart from this essential skills module, Lap VR has a procedural training module for Laparoscopic Cholecystectomy and also for various procedures like tubal occlusion and salpingo oopherectomy in the field of gynecology [22].
3. How effective is VR for Surgery? Virtual reality technology, after its success in flight simulators, the application of this technology was tried in surgery. In the previous section, we discussed about a number of virtual-reality 29
based laparoscopic training simulators which were developed to train users on laparoscopic skills. But the biggest question is, whether these virtual reality based simulators are effective in training for laparoscopic skills. Therefore, it is necessary to examine the impact of virtual reality based laparoscopic simulators on the improvement of specific laparoscopic skills in the real world. In other words, there is a need to validate the transfer of laparoscopic skills from the virtual reality setup to the real world.
a.
Validation Studies on the VR Simulators
The conventional methods of laparoscopic training before the introduction of VR technology was by using animal models and by using laparoscopic box trainers. Therefore, the initial validation studies were focused on proving the simulator is better than conventional methods.
Grantcharov et.al.[24] showed that the results of training on animal model is comparable to the results obtained by training on virtual models. Therefore, computer models are equally effective to physical models in training for laparoscopy. But, the physical animal models are very expensive compared to the computer models.
Torkington et.al.[25] compared the conventional box trainers and the VR trainers in transferring the skills to a real laparoscopic task, and found that results achieved with the conventional box trainers and the VR trainers were similar, and both of them transferred skills to a real laparoscopic task. But, there is no objective assessment of skill acquisition with the use of laparoscopic box trainers. Thus, virtual reality training scored over physical animal models and the conventional box trainers for laparoscopic skills training.
Gallaghar et.al.[26] found that virtual reality training helps the novices overcome the problem of “fulcrum effect” in laparoscopic surgery and concluded that VR training offers a safe and realistic learning environment which can quantify the skill level of the users. Gallaghar et.al.[27] also proved the construct validity of the simulator by proving that the simulator is able to differentiate between subjects with varying skill levels. Experienced surgeons performed the tasks much faster and with fewer errors as compared to the novices. 30
i. MIST-VR Investigations above show that the virtual reality simulator helps to acquire laparoscopic psychomotor skills and is a reliable tool for the assessment of the same. Then studies were carried to find out if MIST-VR helps in improving surgical performance of actual laparoscopic procedures.
Ahlberg et.al.[28] studied if preoperative training on MIST-VR helps in performing laparoscopic appendectomy on a pig. The results were negative; there was no significant improvement in the performance of inexperienced surgeons after training on the MIST-VR.
Seymour et.al.[29] demonstrated that the virtual reality training improves the operating room performance in laparoscopic cholecystectomy. VR trained group completed the procedure significantly faster, with fewer number of errors when compared to the control group with no prior training. In this study, the VR trained group was trained until a desired level proficiency was achieved.
McClusky et.al.[30] reiterated that criteria based training (to a certain skill proficiency) on MIST-VR is better than standard training. In his study, the group which had criteria based training performed significantly better in the laparoscopic cholecystectomy than the control group (which had standard training).
Grantcharov et.al.[31] studied the impact of virtual reality surgical simulation on the improvement of psychomotor skills for performing laparoscopic cholecystectomy, by using MIST-VR and found that surgeons who received VR training on MIST-VR performed laparoscopic cholecystectomy significantly faster than the surgeons who had no VR training.
Thus, from all the above studies, it is clear that the Virtual reality based simulator MIST-VR, not only has positive impact on laparoscopic psychomotor skills, but also helps to deliver better operating room performances.
31
ii. LapSIM LapSim is another virtual reality simulator that has been widely studied. The results obtained by these studies proved that virtual reality trainers transfer laparoscopic skills to the operating room.
Hyltander et.al.[32] showed that the basic psychomotor skills are acquired by novices training on a virtual reality based laparoscopic simulator. The group trained on LapSim performed significantly better, taking lesser time with a lesser error score than the group which had no training on a porcine model.
Expert surgeons performed significantly better than the novices on LapSim. Thus, Duffy et.al.[33] showed that LapSim has construct validity to distinguish between expert and novice surgeons. Van Dongen et.al.[34] and Eriksen et.al.[35] also reiterated the same by their study.
Youngblood et.al.[36] showed that novices trained on virtual-reality simulator LapSim, performed better on a live surgical task on a porcine model than those trained on the conventional box trainer.
Thus, from the studies done on MIST-VR and LapSim, it is proved that:
Virtual reality laparoscopic simulators provide objective assessment of laparoscopic skills.
Virtual reality based laparoscopic simulators not only help novices acquire laparoscopic skills, but also help experts hone their skills.
Virtual reality based laparoscopic simulators are better than conventional methods of training using physical models or using the conventional box trainers.
Virtual reality based laparoscopic simulators helps transferring laparoscopic surgical skills to the operating room thus improving operating room performances.
32
4. Development of Cognitive Skills for Laparoscopic Training The above mentioned simulators train the residents well for the motor skill improvement. However, none of them really address the issue of improving the cognitive ability of the resident being trained for motor skills. The residents would generally get acquainted to identifying the anatomical landmarks by watching more surgeries and training under a surgeon. This would make the learning curve slow and difficult.
Wei Jin et al.[46] developed a system for training which incorporated scenarios by spatiotemporally utilizing videos recorded during actual surgical procedures. This would give the residents a feel for the actual surgical procedure in a controlled virtual environment.
Guerlain et al.[47] came up with the concept of having an entire surgical procedure shown in a sequence of events and then test the effectiveness of the system by evaluating the students using questions assessing perceptual knowledge. Their findings proved that this had a significant increase in scores on questions assessing perceptual knowledge and procedural knowledge, with no corresponding increase for the control group who watched the videos from the same cases but in an unstructured format for the same amount of time. Neither group showed improvement on strategic or declarative knowledge tests.
Kawai T et al.[48] used a 3D reconstruction of the 2D video available from actual surgeries and used them to train the students. They used this technique primarily for minimally invasive procedures. The video was edited stereoscopically at the Waseda University and converted into a streaming 3D video format.
Ruthenbeck et al.[53] developed a VR anatomy puzzle which was a 3D jigsaw puzzle of the skull to facilitate the learning of the anatomy of the skull. A preliminary investigation proved that this was an effective learning tool in comprehending the anatomy of the skull. Pugh et al. [54] conducted a study on Stanford University‟s E-Pelvis Simulator. It is a virtual simulator to learn the anatomy of the female reproductive system. The study concluded that the 33
students in the simulator group learned the anatomy better as compared to the group trained on the traditional manikin system.
The addition of the cognitive skills along with the motor skills in a virtual environment would make a complete simulator that could help in training on all aspects of a surgical procedure.
5. Simulators for the daVinci Surgical Simulator(DVSS) Hayashibe et.al.[37] was the first group to develop a training simulator for the DVSS. They developed a simulator for the preoperative setup of the DVSS. The simulator trains the surgeons to assess the sites for the placement of the trocars. This simulator does not train the surgeons to manipulate the instruments.
a. SurgicalSim Education Platform (STEP) Sim Surgery Inc. developed a commercial product, The SurgicalSim Education Platform (STEP) [38] for robotic surgery training. STEP consists of a basic module for training of hand-eye coordination in robotic surgery and an advanced module for the training of more complex tasks and procedures. The hardware platform of STEP is as shown in figure 22, is a two handed 7 degrees of freedom system.
34
Figure 22: Setup of SurgicalSim Education Platform (STEP)
Halvorsen et.al.[39] proved that Virtual reality training can replace the standard way of training during the early part of training. No information has been published about the skill tasks in both the basic and advanced modules. The input device however does not mimic a real robotic console but uses a free hand movement using a 6 degrees of freedom input device.
b. daVinci Surgical Simulator by the Chinese University of Hong Kong Loi-Wah Sun et.al.[12] from, the Chinese University of Hong Kong, designed and developed a simulator for the daVinci Surgical System. The Phantom Omni device, from Sensable Technologies, was modified with a new electromechanical gripper device to control both arms and instruments in a 3D virtual environment, as shown in figure 23. Motion of the arm was controlled by inverse kinematics based on commonly used Jacobian technique. The basic principle of inverse kinematics is opposite to that of forward kinematics, whereby joint orientation parameters of the robotic arm are incrementally changed from the initial position towards a configuration state that will result in tool tip being located at the desired position.
Figure 23: Setup of the daVinci simulator of Chinese University of Hong Kong. 35
The most important feature of their work is the electromechanical gripper that replaces the phantom stylus. The gripper was used to realistically simulate the opening and closing of the tool tip. The gripper is as shown in figure 24.
Figure 24: Electromechanical gripper.
The gripper consists of two parts, a base part and a movable part. The base part is fixed and has an extension that connects to the stereo plug of the phantom. The base part is connected to the movable part by a ball bearing and axis. There is a cylindrical extrusion that houses a magnet. The movable part has a magnetic sensor and a printed circuit board. The sensor is centered to the rotational axis. When the movable part moves (i.e. when the gripper closes), the magnet rotates about the axis, changing the direction of magnetic field. The magnetic sensor acquires this information and the information is processed in the printed circuit board, translated to angular values and then transmitted to the computer via a USB interface. OpenSceneGraph (OSG) [40], an open source C++ tool kit for 3D graphics, based on OpenGL, was used for rendering. Phantom Omni device was interfaced to the application by using osgHaptics [41], a tool kit that integrates OpenHaptics into OSG. A C++ library was written to communicate the gripper attached to the phantom stylus. No physics engine was integrated in to the application to offer realistic physics simulation.
36
A training scenario program was developed to test the feasibility of simulating a bean drop drill. Users can manipulate a virtual daVinci robot to practice picking up and pasting two spheres and two blocks as shown in figure 25.
Figure 25: Bean drop drill using the daVinci surgical simulator of the Chinese University of Hong Kong
c. Institute for High Dimensional Medical Imaging, The Jikei University School of Medicine, Tokyo, Japan Suzuki et.al.[42] developed a tele-training simulator to enable a surgeon to master and practice the techniques of robotic surgery, the DVSS in particular. This simulator uses the commercial device Phantom as the hardware interface. The movement of each link of the virtual forceps is mapped to the movement of each link of the Phantom. A sphere filled model well suited for realtime deformation is used to train users with surgical maneuvers such as grasping, pushing and ablation.
The sphere filled model is a soft-tissue model, consisting of a group of small rigid spheres of the same radius and a triangular mesh at the surface. It was constructed such that the inner portion of any organ was completely filled with layers of spheres. A patient organ model was reconstructed in to the sphere filled model as shown in figure 26.
37
Figure 26: Patients organ model on the left and its corresponding sphere filled model on the right.
Gall bladder was modeled as yellow colored spheres of radius 2mm, while the liver was modeled as red spheres of 6mm radius. This model trains the user to perform cholecystectomy (removal of gall bladder). Grasping, pushing and detachment were the basic tasks trained in the model.
Grasping was accomplished by an algorithm that identifies the nearest sphere in contact with forceps, and makes that sphere follow the tool tip. Surrounding spheres that make up the organ surface was made to move in a direction towards the grasp point.
Detachment was accomplished by another algorithm, which first identifies the spheres that are in contact region of the gall bladder and liver. Then the algorithm checks if the left forceps grasp the gall bladder and the right forceps grasp the liver. If so, detachment of the spheres in the contact region happens, and the gall bladder and liver models separate.
The capability of the simulator was extended to make it a tele-surgical simulator as shown in figure 27.
38
Figure 27: Tele-surgical simulator for the DVSS, by Jikei University School of Medicine.
d. The dV Trainer by Mimic Technologies Mimic Technologies developed a trainer for the DVSS called the dV Trainer [44]. The basic setup is as shown in figure 28. It is a custom haptic device for training laparoscopic procedures which replicates the DVSS‟s surgical console. It is currently under research and development. The master of this system uses a novel patented string based system instead of a kinematic system used in the actual system.
The training modes are divided into two categories:
System Training
Skills Training
System Training: In systems training the surgeon gets accustomed to the master controller of the DVSS. The various tasks that are used to train for the same are:
1. Surgeon Console Awareness: The surgeon is made comfortable with the various controls of the master console. 2. EndoWrist Manipulation: The various motions possible using the EndoWrist are taught to the surgeon. 39
3. Camera and Clutching: The use of camera manipulation and the clutching pedals are trained for. 4. System Assessment
Skills Training: The basic skills needed to perform a robotic laparoscopic surgery are developed by the skills training. The various tasks being:
1. Needle Exchange: Transferring the needle from one tool to another. 2. Needle Driving: Putting the needle through a tissue. 3. Knot Tying: This essentially deals with being able to tie a knot using the needle thread and the two tools in the environment. 4. Suturing: This teaches the trainees to tie suture the organs together at the end of the procedure.
Figure 28: The dV trainer developed by Mimic Technologies
Lendvay et al. [52] performed a randomized blinded pilot study on the dV trainer. The subjects performed a ring transfer module on the DVSS and on the dV trainer. It was concluded that an offline robotic trainer is comparable to the dry lab robotic skills station, and this trainer was able to discriminate between novices and experts of robotic surgery, thereby meeting all criteria for being an effective simulator. The evaluation of the surveys showed that there was a role of simulation in surgery and 93% believed that an offline trainer would be useful to train people to use the DVSS. 40
6. Redesign of the Robotic Surgical Simulator (RoSS) for the DVSS From the validation studies performed on LapSim and MIST-VR, it is clear that virtual reality technology improves laparoscopic psychomotor skills. Also, from the studies made on the virtual reality based laparoscopic simulators, it is evident that a basic training module for a laparoscopic surgery must impart training for specific skills like camera navigation, instrument navigation, grasping, bimanual coordination, stretching, cutting, and clip applying.
The first design of the RoSS system [50] consisted of two psychomotor skill tasks. The two tasks were ball drop and the needle capping task.
Ball Drop: The task is as shown in figure 29. The objective of the task is to pick all the four balls one at a time and put in one bowl and then transfer them to the other bowl. This task basically trained for instrument navigation and grasping skills.
Figure 29: Ball Drop
Cylinder Capping: The task is as shown in figure 30. The objective of the task is to pick one cylinder at a time and place it in the pegs on the peg board. This task basically trained for needle capping skills. 41
Figure 30: Cylinder capping
This system was evaluated and studies were performed to check the effectiveness of the system. The studies concluded that the RoSS system built in-house was effective in training. However, the old system did not account for the kinematic accuracy of the surrogate master. Hence, in the new system, it was decided to improve the kinematic accuracy of the system by performing inverse kinematics and enhance the existing psychomotor skill set and building a new module for cognitive skills training.
The psychomotor tasks:
Pick and Place: The objective is to acquire the sphere and place it inside the box. This task was to train instrument navigation, grasping and sense of depth perception.
Tissue Traversal: The objective of this task is to traverse along the length of the cylinder to reach from one end to another. This trains grasping skills using the forceps and being able to walk down along the length of the tissue.
Stretching and Clipping: This involves the pulling of the cylinder along its length and clipping at the desired point. This trains for stretching and clipping the tissue without causing dissection or losing contact. 42
The cognitive skills sets consist of the following:
Anatomy Identification: This involves a series of anatomical tests. In each test a frame from the actual surgery is shown to the trainee and trainee is required to identify the different tissues and organs in that scene.
Bleeding Control: This involves a series of different scenarios involving bleeding and also some where there is no bleeding occurring. The trainee has to identify the bleeding correctly, and take remedial actions by choosing a tool. Since, the old RoSS system didn‟t account for the actual kinematics of the DVSS the old RoSS console was redesigned to improve the flexibility and ergonomics. Inverse Kinematics was implemented to ensure that the system was kinematically correct.
Design parameters required for the final system:
Hardware system must simulate the daVinci console and also have the various functionalities like the pedals.
Motion of the virtual tool using the surrogate master controller should be similar to the motion of the slave of the DVSS using the master controller of the same.
It was proposed to develop a simulator satisfying these design requirements, increasing the number of psychomotor tasks and adding a module for the cognitive skills development.
Development of RoSS simulator is a large project currently under way at University at Buffalo, VR lab in collaboration with Roswell Park Cancer Institute. This project consists of several simulation modules. In this thesis, evaluation has been done for only the cognitive skills set.
43
Chapter 3: Development of Hardware Interface The objective was to develop a setup which is physically identical to that of the DVSS console. Figure 31 shows the set up of the daVinci surgical simulator console where the surgeon sits and performs the robotic surgery viewing the stereoscopic display.
Figure 31: Setup of DVSS
While building the new system, it was critical to ensure that the kinesthetics of working with the DVSS is maintained along with its conformation to the kinematics. The Phantom Omni, a commercial product of Sensable Technologies, was used as the master for the simulator that was built. The Phantom Omni is the most cost-effective haptic device available in the market for users to touch and manipulate virtual objects. Its IEEE1394a fire wire interface ensures quick installation and ease of use. The Phantom Omni is six degree of freedom device which give the position and force feedback [43]. The Phantom Omni was mounted upside down in order to match the kinematic linkages of the DVSS master console. The Omni device has a shoulder and an arm which account for three degrees of freedom. When mounted upside down the arms of the Omni closely resemble the configuration of the arms of the actual system. Furthermore, the devices are mounted at an angle of 35 degrees inwards. This is done to ensure that the surrogate 44
master (Omni device) has a restricted motion on the outside just as in the case of the actual system. Also, the range of motion obtained when the Omni is mounted in this configuration is similar to that obtained in the real master controller. Figure 32 shows the daVinci master and the Phantom Omni mounted upside down at an angle 35 degrees inwards.
Figure 32: daVinci master (left) and Phantom Omni (right) mounted upside down.
1. Design of the Frame The Omni devices are mounted onto a plate that is mounted on the structure as shown in figure 33. The Omni device is tapped in the base to facilitate its mounting on the base plate. Once the device is mounted the plate is attached to the structure.
Figure 33: The Omni mounted on the frame. 45
A monitor was mounted on a fixture such that the positioning of the monitor with respect to that of the Omni devices is the same as that of the DVSS master. To achieve the mounting of the monitor and the Omni device identical to that of the DVSS master, a frame structure was designed which was completely collapsible and flexible for future modifications as shown in figure 34.
Figure 34: Setup of the frame structure
The display monitor is mounted on the slanting frame at the top this ensures the similarity in the way the surgeon looks at the screen. The Omni devices are mounted on the on a base plate which is in turn mounted on the frame either side of the monitor as shown in figure 36. The foot pedals that are placed at the bottom are used for various functions such as clutch, camera, etc.
46
Figure 35: The Simulator
2. Design of the Wrist The stylus of the Omni device is replaced with a custom wrist which gives the feel of working on the EndoWrist of the DVSS. The basic design is as shown in figure 36.
Figure 36: The DVSS EndoWrist (left) and the custom Wrist (right) for Omni
47
The wrist is manufactured using a plastic tube as the outer cover and a stereo jack that fits inside the Omni. It consists of a button press which is which is engaged when the rods are pressed with both the fingers. The button is in turn connected to a stereo jack which completes the circuit when the button is pressed. When the button is pressed, the circuit completes and a one is recorded in the encoder.
48
Chapter 4: Development of Graphical Interface In developing the software interface, the most vital challenge is to match the movement of the virtual tool, to the movement of tool (EndoWrist Instrument) in the DVSS. This was achieved by performing the inverse kinematics to relate the workspaces of the DVSS master and the RoSS master and the workspaces of the slave and the virtual tool respectively.
1. Mapping using Inverse Kinematics The line diagrams of the master of DVSS and the master of RoSS are as shown in figure 37. The master of the DVSS can be split into the arm and the wrist. The arm consists of a shoulder and an elbow. The shoulder has two degrees of freedom and the elbow has one degree of freedom. The EndoWrist has five degrees of freedom including one for the pinch. .
Figure 37: The line diagram of the master controller of DVSS (left) The line diagram of the master controller of RoSS (right)
49
The master of the DVSS controls a highly articulated tool which has seven degrees of freedom. The Omni device can also be divided into two parts the arm and the wrist. The arm consists of the shoulder and an elbow. The shoulder has two degrees of freedom and the elbow has one degree of freedom. The wrist of the Omni consists of 4 degrees of freedom including one for the button press (pinch).
Based on the line diagrams of the two systems the Denavit-Hartenberg (DH) [45] parameters for the two systems are evaluated using the algorithm (See Appendix).
The DH parameters for the DVSS master and the RoSS master are as shown in Table 1 and Table 2 respectively.
Link
θ
d
a
α
1
θ1
d1
0
-pi/2
2
θ2
0
L2
0
3
θ3
0
L3
-pi/2
4
θ4
d4
0
pi/2
5
θ5
0
L5
-pi/2
Table 1: DH Parameters of DVSS master
50
Link
θ
d
a
α
1
θ1
d1
0
-pi/2
2
θ2
0
L2
0
3
θ3
0
0
-pi/2
4
θ4
d4
0
pi/2
5
θ5
0
0
-pi/2
Table 2: DH Parameters of RoSS master
Based on the DH parameters the transformation matrix for each link is obtained using
cos i sin i i 1 Ti 0 0
- sin i cos i - cos i cos i sin i 0
sin i sin i sin i cos i cos i 0
a i cos i a i sin i di 1
and the final transformation for all the links is obtained as follows.
T = T1 * T2 * T3 * T4 * T5
Once the final transformation matrix is obtained inverse kinematics algorithm (See Appendix) is applied to obtain the joint angles of the slave for a set of angles moved by the master.
51
2. Omni Interface Open Haptics API [43], an API distributed with the device, was used for interfacing the device. The API primarily is written in C++ and has libraries namely HD and HL for handling the device. These libraries enable the user to read the raw encoder values for the joint angles, gimbal angles, forces at the joints and the device position. The API also enables the use of multiple Omni devices hence, two devices are used.
3. 3D Graphics Renderer OpenGL and GLUT are used for the 3D graphic rendering. It also enables the use of stereographic rendering which is used as a feature in RoSS. All the graphics objects including the tools were modeled in SolidEdge and imported into OpenGL.
4. Features The next version of RoSS incorporates almost all the features of the DVSS.
Stereo Vision: The DVSS uses a stereographic display. The surgeon on the master controller sees everything in 3D. Hence, the 3D stereoscopic vision is included which can be achieved by using 3D glasses.
Clutch Pedal: This is a very important function in the DVSS because whenever the surgeon reaches an unfavorable position he/she can use this function and get the tool back to the favorable position and continue the operation. The clutch functions as follows: as soon as the clutch is pressed the slave is disengaged from the master and does not move until the clutch is released. Just like in the actual system the clutch function in RoSS is engaged by using foot pedal named clutch.
52
Camera Pedal: The camera function essentially enables the navigation through the body. When the camera pedal is pressed the tools don‟t move. The camera arms move as the master is moved. This enables the surgeon to access all sections inside the body. The camera function in RoSS is engaged by using the pedal named camera.
Shadows: One of the important features incorporated in the RoSS system is the use of shadows. Shadow helps in understanding the depth in a 3D world.
Reset/Restart: During training the trainee can reset any trial or restart the whole experiment to practice further and become more proficient at using the robot.
Error Metrics: The different error metrics such as collision between the tools, the unwanted collisions between the objects in the scene and the tool and the time required to complete the task are automatically recorded.
The complete assembled RoSS system is as shown in figure 38.
Figure 38: The complete RoSS system. 53
5. Development of Skill Sets a. Development of Psychomotor Skill Sets The psychomotor skill sets were developed using OpenGL and GLUT as the graphics renderers and open haptics API for the interface of the Omni with the system. The collision detection for the graphics shapes is done using hierarchical sphere-sphere collision detection algorithm. The trainee can use the different pedal functions to complete the tasks.
1. Pick and Place The general idea of the pick and place task is to acquire the sphere in the scene and place it inside the box. When the tip of the tool is near the sphere, by gripping the tool the ball can be acquired. Using the clutch and the camera functions the user can reach the box with the sphere inside the jaws of the tool. As soon as the sphere is completely inside the box, the box turns green. At this moment the sphere is released. The same is repeated multiple times and the performance is evaluated. The task being performed is shown in figure 39.
Figure 39: Pick and Place .
54
2. Traversal The general idea of the traversal task is to move along the length of the cylinder successively. The cylinder is divided into three sections. The first section is purple in color at the start. As soon as the first part is gripped the second section is highlighted with green color. This section is gripped with the other tool. Once the second is gripped the third section highlights in green color and the third section is gripped by releasing the first section and still gripping the second section. The same is repeated multiple times and the performance is evaluated. The task being performed is as shown in figure 40.
Figure 40: Traversal
3. Stretching and Clipping The general idea of the stretching and clipping task is to stretch the cylinder using by holding the sphere and moving along the length of the cylinder. A blue band would appear when the cylinder has been pulled to the required length. Using the second tool the cylinder is clipped at the blue band. The same is repeated multiple times and the performance is evaluated. The task being performed is as shown in figure 41.
55
Figure 41: Stretching and Clipping
b. Development of Cognitive Skill Sets The cognitive skill sets were developed using OpenGL and GLUT as the graphics renderers and open haptics API for the interface of the Omni with the system. Videos from actual surgical procedures are obtained. The video feed from the surgery is obtained in a .mpg format. They are converted into .avi formats using the software called Any Video Converter (free software [49]). The .avi file thus obtained is then edited to obtain the video in the form of multiple frames. A software called avi4bmp (free software [49]) is used to convert the video into frames. The frames thus obtained are in the .bmp format. These frames are converted from the .bmp format to the .tga format using the software called IrfanView (free software [49]). These frames are read into the system and stored in the texture memory of Opengl. These textures are rendered on a plane at the rate of 40 textures per second. The trainee would interact with this video and perform the requisite task. The interaction is obtained with hierarchical sphere-sphere collision detection. The object of interest has a collision shape associated with it and if contact is made then the requisite action is performed. The depth cues in the system are provided using green markers. Hence, every time the tool tip touches the surface of the video a green marker would appear at that point. The trainee can use the different pedal functions to complete the tasks.
56
i. Anatomy Identification: To perform surgery using a minimally invasive procedure, the surgeon has to have a good understanding of the various organs from the perspective of the internal view of the pelvic region provided by the camera. This would help in ensuring that the correct tissues are cut. Also, it is important to know all the surrounding regions thoroughly.
A skill set with anatomy identification system was build to help teach medical personnel the anatomy from the camera’s point view as opposed to that of the images provided in text books. This system is composed of a series of anatomical tests with various landmarks. The trainee is presented with one scenario at a time wherein the trainee needs to identify the different organs and tissues in the scene. The trainee can verify the answer by selecting the explanation option. The critical aspect of this task is to evaluate how many anatomical parts are identified correctly and the time taken to identify them. The task being performed is as shown in figure 42. (i) Shows the view of the right side of the posterior peritoneum and the ureter, sigmoid and external illiac arteries are seen. (ii) Shows the Lateral Pelvic space on the left of posterior peritoneal edge. The ureter and the superior vesicle artery are seen. (iii) Shows the view of the female pelvic region with the sigmoid and the uterus seen. (iv) At the beginning of the ureteral dissection the sigmoid the ovarian pedicle and the uterus are seen. (v) It is the view of the paravesical space and the uterine artery and sigmoid are seen. (vi) It is the view of the para -vesical space and the ureter and the uterine artery are visible.
57
Figure 42: Anatomy Identification
ii. Bleeding Control The general idea of bleeding control is to identify the area on the video where the bleeding is occurring. Once the exact position of the bleeding is known, using the tool the bleeding is controlled by cauterizing the region. The clutch and camera function is used to navigate through the scene and reach the appropriate position. Once the position is reached the tool is gripped and the bleeding is clipped or cauterized thereby stopping the bleeding. The task consists of multiple scenes. The trainee is presented with one scenario at a time and the trainee has to recognize the location of the bleeding. In spaced in this module are situation where no bleeding is occurring. In such a case the trainee needs to identify that there is no bleeding. If the time taken is too high the trainee can look at the solution and try the task again. The task being performed is as shown in figure 43.
58
Figure 43: Bleeding Control: The pictures on the left show two different scenarios of occurrence of bleeding and the use of virtual tool to cauterize the region, the pictures on the right show two scenarios where no bleeding in the system.
59
Chapter 5: Experimental Details The previous version of RoSS when evaluated showed that it was an effective training tool [50, 51]. Also, it has been shown that using video clips in teaching the medical students along with the textbook material makes them better equipped to perform surgeries in future [47]. The aim of the present experimental study was to evaluate if interaction with the videos would make them better robotic surgeons. To achieve this, a pelvic anatomy syllabus and a final test were devised. The syllabus consisted of line diagrams and color pictures of the human anatomy. The test consisted of questions that tested their knowledge of the pelvic anatomy.
Guerlain S. et al.[47] investigated if using videos from actual surgical procedures improves the skills of students. They used an ordered set of video clips that showed the main steps of a laparoscopic procedure. Each step was shown several times and from a different surgery. They proved that these ordered perceptual learning modules had a significant increase in scores on questions assessing perceptual knowledge and procedural knowledge as compared to the control group.
1. Experimental Protocol The total number of subjects used for this study was 10. They were medical students from the University at Buffalo and residents from the Roswell Park Cancer Institute. The subjects were divided into two groups. The two groups were: Group I: Provided with textbook version of anatomy without video on RoSS. Group II: Provided with textbook version of anatomy and video on RoSS. All the participants were required to complete a pre-task survey to obtain information about their medical background. Subsequently, all the subjects were given directions about the entire procedure.
60
The first group was given a booklet containing text, line diagrams and colored photographs of the anatomy in the pelvic region. They were given one day to review the entire material. They took a test where in they were required to identify certain anatomical landmarks and mark them on the photographs which were screen shots from an actual laparoscopic procedure. They were required to identify specific organs and vessels in this module. The time allotted to complete the entire test was 150 seconds. They were asked to complete a post test survey. The second group was given a booklet containing text and line diagrams of the anatomy in the pelvic region. They were given one day to review the material. After this, they were trained on the RoSS system using the cognitive skill sets. They took a test where in they were required to identify certain anatomical landmarks and mark them on the photograph which was a screen shot from an actual laparoscopic procedure. This test was identical to the one given to Group I. They were required to identify specific organs/vessels in this module. The time allotted to complete the entire test was 150 seconds. They were asked to complete a post test survey.
2. Experimental Results The test given to the participants tested them on their knowledge of anatomy. The same test was given to both the groups to study the effect of the RoSS trainer. The test given consisted of five different screen shots of anatomical significance in while performing surgery in the pelvic region. Each screen shot was accompanied with a question where in the participant was expected to identify the anatomical structure of the organ or vessel mentioned in the question. Subsequently, the participant had to outline the correct structure in the photograph. They were required to complete all the five questions in less than 150 seconds. The test results were evaluated for: (i) total time taken to complete the test by both the groups (ii) number or correct answers (iii) the number of mistakes/error in identifying the various anatomical landmarks. The means for each of these parameters is as listed in the table 3. 61
Time Groups
Errors Standard
Mean(Seconds)
Correct Answers
Standard Mean
Deviation
Standard Mean(out of 5)
Deviation
Deviation
Group I
142.8
10.733
1.7
0.44
2.9
0.223
Group II
118.4
19.034
0.4
0.5
4.2
0.758
Table 3: Means of the various parameters Mean time taken for the group trained on cognitive skills on RoSS along with the textbook material was 118.4 seconds which was considerably less than the time taken (142.8 seconds) by the group that used only textbook material and did not undergo any additional training. Figure 44 shows the comparison of the mean times taken by both the groups.
Figure 44: The box plots for the mean times for completion of the test Hypothesis testing for the mean times of completion of the test. H o:
=
(No effect of RoSS training)
H 1:
≠
(RoSS training has effect)
62
Null Hypothesis: RoSS training has no effect on the mean times for completion of the test. ANOVA was performed using a statistical analysis software package Minitab, using a general linear model. The result was F = 6.23, P = 0.037. Therefore with 96.3% confidence, the null hypothesis was rejected. In other words, statistically it was validated with 96.3 % confidence that training on RoSS was responsible for the decrease in the mean time taken by the group trained on the cognitive skill sets on RoSS.
The errors signify mistakes in the identification by of the anatomy. Mean number of errors committed by the group trained on RoSS was 0.4 out of 5 while the group that did not undergo training on RoSS committed 1.7 out of 5. Figure 45 shows the comparison of the mean errors committed by both the groups.
Figure 45: The box plots for the mean number of errors for the test
63
Hypothesis testing for the mean number of errors in the test. H o:
=
(No effect of RoSS training)
H 1:
≠
(RoSS training has effect)
Null Hypothesis: RoSS training has no effect on the mean number of errors in the test. ANOVA was performed using a statistical analysis software package Minitab, using a general linear model. The result was F = 16, P = 0.004. Therefore with 99.6% confidence, the null hypothesis was rejected. In other words, statistically it was validated with 99.6 % confidence that training on RoSS was responsible for the decrease in the mean number of errors by the group trained on the cognitive skill sets on RoSS.
The number of correct answers was the total number of correct answers given by the subjects. The mean number of correct answers given by the group trained on RoSS was 4.2 out of 5 while the group that did not undergo any training gave 2.9 out of 5. Figure 46 shows the comparison of the mean number of correct answers given by both the groups.
Figure 46: The box plots for the mean number of correct answers for the test
64
Hypothesis testing for the mean number of correct answers given in the test. H o:
=
(No effect of RoSS training)
H 1:
≠
(RoSS training has effect)
Null Hypothesis: RoSS training has no effect on the mean number of correct answers in the test. ANOVA was performed using a statistical analysis software package Minitab, using a general linear model. The result was F = 13.52, P = 0.006. Therefore with 99.4% confidence, the null hypothesis was rejected. In other words, statistically it was validated with 99.4 % confidence that training on RoSS was responsible for the increase in the mean number of correct answers by the group trained on the cognitive skill sets on RoSS.
3. Results of the Post Task Survey The subjects were asked to evaluate the skill sets as learning tools and also the general rating for the RoSS setup. The subjects found the anatomy module helpful in enabling them better understand the anatomy. The rating for the module was 3.6 out of 5. Figure 47 depicts the general rating by the subjects. (See Appendix for the surveys)
Figure 47: The rating of the anatomy module.
65
The subjects found the bleeding module helpful in helping them get better equipped with the actual bleeding scenarios that could occur while performing the surgeries. The rating for the module was 3.6 out of 5. Figure 48 depicts the general rating by the subjects.
Figure 48: The rating of the anatomy module.
The subjects found the overall working on RoSS as a comfortable experience. The rating for the overall comfort on working with the system was 3.6 out of 5. Figure 49 depicts the general rating by the subjects.
Figure 49: The overall rating of the RoSS syste 66
Chapter 6: Conclusions and Future Work The RoSS system has a set of psychomotor skills and a set of cognitive skills. The additional components added to this system include:
a more compact structure that resembles the DVSS more closely
a custom wrist which imitates the wrist of the DVSS
the pedal functions.
The psychomotor skill sets train for hand eye coordination and the cognitive skill sets provide the surgeons with specific skills for performing specific laparoscopic surgery. The results of the studied concluded that:
there is a reduction in the time taken to identify the various anatomical landmarks.
there is a reduction in the number of errors after the training
there is an increase in the number of correct answers after the training.
Also, after the study the subjects felt,
The RoSS system is a comfortable system to work with.
The anatomy module helps in improving the anatomy identification skills.
The bleeding module helps in making them better equipped with the real surgical scenarios.
Therefore, an effective simulator was built with a set of basic psychomotor skill set and a basic cognitive skill set which would improve the performance of the subjects on the real system. The psychomotor skill set trains for the basic hand eye coordination and the cognitive skill sets improve the anatomy identification skills and the bleeding module improves the bleeding identification skills. This leaves us with a wide scope of future work to be done. 67
The following is the immediate work that can be performed.
The wrist designed needs an extra degree of freedom so that it does not entangle the wrist of the surgeon.
The electro-cautery function to be incorporated in the pedals.
Developing a full set of anatomical identification skill sets for different procedures.
Developing a skill set to train for suturing.
The study would be conducted as follows.
All the participants would take a pre-test to assess their anatomical knowledge.
Each one will fill out a survey.
Every individual will be given the line diagrams and text for review
They would be tested on the study material to evaluate their improvement
After this, they would be divided into two groups and one group will get the pictures and photographs of the anatomy and the other group would be trained on RoSS.
Every person from both the groups would answer a final test to evaluate the performance.
The group which trained on RoSS would answer a post-task survey which will evaluated for the effectiveness of the system.
68
Appendix 1) The algorithm to find the DH parameters •
Step 1: Assign a coordinate frame L0 to the robot base.
•
Step 2: Align zk with the axis of joint k + 1.
•
Step 3: Locate the origin of L k at the intersection of zk and zk-1 .When there is no intersection, use the intersection of zk with a common normal between zk and zk-1
•
Step 4: Select xk to be orthogonal to zk and zk- 1. lf zk and zk-1 are parallel, point xk away from zk- 1 .
•
Step 5: Select yk to form a right handed orthonormal coordinate frame.
where, θk is the angle of rotation from xk-1 to xk measured about zk- 1. dk is the distance measured along zk- 1 ak is the distance measured along xk. αk is the angle of rotation from zk- 1 to zk about xk
2) The inverse kinematics algorithm •
Calculate the difference between the goal position and the actual position of the end-effector
•
Calculate the Jacobian matrix using the current joint angles
•
Calculate the pseudo-inverse of the Jacobian
•
Determine the error of the pseudo-inverse error
•
If error > e then dX = dX / 2 restart at step 4
•
Calculate the updated values for the joint orientations and use these as the new current values. Check the bounds for theta values.
•
Using forward kinematics determine whether the new joint orientations position the end-effector close enough to the desired absolute location. If the solution is adequate then terminate the algorithm otherwise go back to step 1
69
3) Pre-Task Survey
1) Current level of training/education: Medical Student (Year): 1
2
Resident (Year): 1
3
2
3
4
2) Current Institution: UB:_______ Other:________________(specify) 3) What level of anatomy training did you have? o Lectures o Lectures + Cadaver Dissection o Dissection + OR Live anatomy 4) Have you ever observed a surgical pelvic anatomy case? o Yes If „yes‟, how many cases? o No ___less than 5 ___greater than 5 5) Did you just observe or assist the case? o Observe o Assist o Neither of the above 6) What would be your preferred way of studying/learning pelvic anatomy? o Lectures o Cadaveric Dissection o Live OR 7) Have you ever sat on a virtual reality simulator? Yes No If „yes‟ which one? ______________________________ Where? _____________________________ How long? ____________ 8) Have you ever used a surgical robot? Yes No If „yes‟ which one? Aesop___ Hermes___ Zeus___ daVinci___ Other________ 9) Is there a „dry‟ lab setting available for training? Yes No if „yes‟, what equipment is available? ____Pelvic Trainers ____Virtual Reality Robot ____Training Robot 70
____Videos 10) Is there a wet laboratory setting available for training? Yes No if „yes‟ what specimens are used? ____Animal ____Cadaveric 11) Do you feel that taking a few courses would be sufficient to incorporate knowledge of MIS into your medical education? Yes No If „yes‟ what type?
For how long?
___didactic
___weeks
___hands-on
___months
4) Post-Task Survey Initials: For the questions with ratings: 1 – Poor 2 – Unsatisfactory 3 – OK 4 – Satisfactory 5 – Excellent
Dominant hand:
1) Did you experience any discomfort in your hands, elbows or shoulders while working on RoSS? Yes No If „yes‟ explain what? ______________________________________________________ 2) Rate the comfort of the seating position. 1 2 3 4 5 3) Did you experience any discomfort in your back or neck while working on RoSS? Yes No If „yes‟ explain what? ______________________________________________________ 4) Were you able to access all positions on screen using the tools? Yes No If no, explain difficulty: ____________________________________________________
71
5) Were you comfortable using the pedals? Yes No If no, explain difficulty: ____________________________________________________ 6) Rate the comfort of working on RoSS. 1 2 3 4 5 Anatomy Module: 7) How similar is anatomy in the operative field to what you have previously learnt? Not Similar Somewhat Similar Similar Same 8) Were you able to extend your prior textbook knowledge to the operative field? Yes No 9) Do you think this training in conjunction with the textbook knowledge help in being better equipped to real surgical scenarios? Yes No Comment: ___________________________________________ 10) Do you think the interaction with a virtual simulator in your education is necessary while identifying the pelvic anatomy? Yes No 11) Rate the module as a learning tool? 1 2 3 4 5 Bleeding Module: 12) Rate the difficulty in identifying the bleeding. Very Difficult Difficult Normal
Easy
Very easy
13) Rate the similarity of this module to when you are watching a live surgery in the OR. 1 2 3 4 5 Not Applicable 14) Do you think the addition of this module would help you further in learning pelvic anatomy in addition to watching live surgeries? Yes No Not Applicable 15) Rate the module as a learning tool? 1 2 3 4 5 16) Do you think the time to perform the tasks was sufficient? Yes No If „no‟ how much time is sufficient? _____minutes
72
References [1] http://www.roboticurology.com/Minimally_Invasive_Robotic_Surgery/Robotic_ Surgery_ 1.jpg [2] Satava RM, Bowersox JC, Mack M, et al., “Robotic surgery: state of the art and future trends”, Contemp Surg. 2001;57:489-499. [3] Prasad SM, Ducko CT, Stephenson ER, et al., “Prospective clinical trial of robotically assisted endoscopic coronary grafting with 1 year follow-up”, Ann Surg. 2001;233:725–732. [4] Sackier JM, Wang Y. “Robotically assisted laparoscopic surgery. From concept to development” , Surg Endosc 1994;8:63–6. [5] Jacobs LK, Shayani V, Sackier JM, “Determination of the learning curve of the AESOP robot”, Surg Endosc 1997;11:54 –5. [6] Schurr MO, Arezzo A, Neisius B, Rininsland H, Hilzinger HU, Dorn J, et al., “Trocar and instrument positioning system TISKA. An assist device for endoscopi c solo surgery”, Surg Endosc 1999;13:528–31. [7] Buess GF, Arezzo A, SchurrMO, Ulmer F, Fisher H, Gumb L, et al., “A new remotecontrolled endoscope positioning system for endoscopic solo surgery”, Surg Endosc 2000;14:395 –9. [8] Schurr M, Arezzo A, Buess GF., “Robotics and systems technology for advanced endoscopic procedures : experiences in general surgery”, Eur J Cardiothora c Surg 1999;16 Suppl 2:S97– 105.
73
[9] Schurr MO, Buess G, Neisius B, Voges U., “Robotics and telemanipulation technologies for endoscopic surgery. A review of the ARTEMIS project. Advanced Robotic Telemanipulator for Minimally Invasive Surgery [see comments]”, Surg Endosc 2000;14:375–81.
[10] http://trueforce.com/Medical_Robotics/Medical_Robotics_Companies/zeus.htm
[11] http://www.intuitivesurgical.com/. [12] Loi-Wah Sun, Frederick Van Meer, Yan Bailly, and Chung Kwong Yeung, “Design and Development of a Da Vinci Surgical System Simulator”, Proceedings of the 2007 IEEE International Conference on Mechatronics and Automation, August 5 - 8, 2007, Harbin, China.
[13] I. A. M. J. Broeders and J. P. Ruurda, "Robotics in laparoscopic surgery: current status and future perspectives," Scand J Gastroentero, vol. Suppl 236, pp. 76-80, 2002.
[14] N. E. Seymour, A. G. Gallagher, S. A. Rance, M. K. O' Brien, U. K. Bansal, and D. K. e. a. Anderson, "Virtual reality training improves operating room performances: results of a randomized, double-blinded study," Ann. Surg., vol. 236, pp. 458-63, 2002. [15] Justin M. Albani, David I. Lee. “ Virtual Reality-Assisted Robotic Surgery Simulation”, Journal of Endourology. 2007, 21(3): 285-287.
[16] J. K. Salisbury, "The heart of microsurgery," Mechanical Engineering Magazine ASME Int'l, vol. 120, pp. 47-51, 1998.
[17] Simulaids®, http://www.simulaids.com/.
[18] S. Cotin, H. Delingette, and N. Ayache, "A hybrid elastic model allowing realtime cutting, deformations and force-feedback for surgery training and simulation," The Visual Computer, vol. 16, pp. 437–52, 2000. 74
[19] S. Pieper, S. Delp, J. Rosen, and S. S. Fisher, "A virtual environment system for simulation of leg surgery," In Proc. of S.P.I.E.: The Int. Soc. for Optical Engineering, Stereoscopic Displays & Applications II, vol. 1457, pp. 188-97, 1991.
[20] C. Basdogen, C. H. Ho, and M. A. Srinivasan, "Virtual environments for medical training: graphical and haptic simulation of laparoscopic common bile duct exploration," IEEE/ASME Transactions on Mechatronics, vol. 6, pp. 269-85, 2001.
[21]http://www.mentice.com/
[22] Simbionix®. http://www.immersion.com/medical/.
[23] Surgical science, Lapsim, Immersion Medical. http://www.surgicalscience.com/
[24] T.P. Grantcharov, J. Rosenberg, E. Pahle and P. Funch-Jensen, “ Virtual reality computer simulation- An objective method for evaluation of laparoscopic surgical skills”, Surg Endosc. 2001 Mar:15(3): 242-4. [25] J. Torkington, S. Smith, T. Brown, B.Rees, and A. Darzi, “Skill Transfer From A Virtual Reality Simulator to a Real Laparoscopic Task”, 2000 Annual Scientific Meeting of the Society of American Gastrointestinal Endoscopic Surgeons (SAGES). [26] Gallagher AG, McClure N, McGuigan J, Crothers I, Browning J, “Virtual reality training in laparoscopic surgery: a preliminary assessment of minimally invasive surgical trainer virtual reality (MIST VR)”, Endoscopy 1999 May; 31(4): 310-3. [27] Gallagher AG, Richie K, McClure N, McGuigan J,” Objective psychomotor skills assessment of experienced, junior, and novice laparoscopists with virtual reality”, World J Surg. 2001 Nov; 25(11): 1478-83. 75
[28] Ahlberg G, Heikkinen T, Iselius L, Leijonmarck CE, Rutqvist J, Arvidsson D, “Does training in a virtual reality simulator improve surgical performance?”, Surg Endosc. 2002 Jan;16(1):126-9.
[29] Seymour NE, Gallagher AG, Roman SA, O'Brien MK, Bansal VK, Andersen DK, Satava RM, “Virtual reality training improves operating room performance: results of a randomized, double blinded study”, Ann Surg. 2002 Oct; 236(4): 458-63; discussion 463-4.
[30] DA McClusky III, AG Gallagher, EM Ritter, AB Lederman, KR Van Sickle, M Baghai, CD Smith, “Virtual reality training improves junior residents‟ operating room performance: Results of a
prospective, randomized,
double-blinded
study
of the
complete laparoscopic
cholecystectomy”, ACS 2004.
[31] Grantcharov TP, Kristiansen VB, Bendix J, Bardram L, Rosenberg J, Funch- Jensen P, “Randomized clinical trial of virtual reality simulation for laparoscopic skills training”, Br J Surg. 2004 Feb;91(2):146-50. [32] Hyltander A, Liljegren E, Rhodin PH, Lonroth H, “The transfer of basic skills learned in a laparoscopic simulator to the operating room”, Surg Endosc. 2002 Sep;16(9):1324-8. [33] Duffy AJ, Hogle NJ, McCarthy H, Lew JI, Egan A, Christos P, Fowler DL, “Construct validity for the LAPSIM laparoscopic surgical simulator”, Surg Endosc. 2005 Mar; 19(3):401-5. [34] van Dongen KW, Tournoij E, van der Zee DC, Schijven MP, Broeders IA, “Construct validity of the LapSim: Can the LapSim virtual reality simulator distinguish between novices and experts?”, Surg Endosc. 2007 Feb 9.
[35] Eriksen JR, Grantcharov T, “Objective assessment of laparoscopic skills using a virtual reality stimulator”, Surg Endosc. 2005 Sep;19(9):1216-9.
76
[36] Youngblood PL, Srivastava S, Curet M, Heinrichs WL, Dev P, Wren SM, “Comparison of training on two laparoscopic simulators and assessment of skills transfer to surgical performance”, J Am Coll Surg. 2005 Apr; 200(4):546-51.
[37] M. Hayashibe, N. Suzuki, M. Hashizume, Y. Kakeji, K. Konishih, S. Suzuki, and A. Hattori, "Preoperative planning system for surgical robotics setup with kinematics and haptics," Int. J Medical Robotics & Computer Assisted Surgery, vol. 1, pp. 76-85, 2005.
[38] SurgicalSim®, http://www.simsurgery.no/robotic_surgery_simulator.htm [39] Halvorsen, F., Elle, O., Dalinin, V., Mørk, B, Sørhus, V., Røtnes, J, and Fosse, E, “Virtual reality simulator training equals mechanical robotic training in improving robot-assisted basic suturing skills”, Journal of Surgical Endoscopy, vol. 20, pp.1565- 1569, 2006.
[40] http://www.openscenegraph.org/.
[41] http://www.vrlab.umu.se/research/osgHaptics/ [42] S Suzuki, N Suzuki, A Hattori, M Hayashibe, K Konishi, Y Kakeji, M Hashizume, “Telesurgery simulation with a patient organ model for robotic surgery training”, The International Journal of Medical Robotics and Computer Assisted Surgery, Volume 1, Issue 4 , Pages80 – 88, 2005.
[43] http://www.sensable.com/ [44] http://www.mimic.ws/MIMIC-dV-Trainer-Flyer-Ltr-Feb08.pdf [45] “Intorduction to Robotics Mechanics and Control” , John Craig , Second Edition, Addison Wesley Publishing Company 1989.
77
[46] Wei Jin, Yi-Je Lim, Tejinder P. Singh, Suvranu De, “Use of Surgical Videos for Realistic Simulation of Surgical Procedures”, Medicine Meets Virtual Reality 14, Volume 119, 2005
[47] Stephanie Guerlain , Kristen Brook Green, Maranda LaFollette, Todd C. Mersch, Brian A. Mitchell, G. Reed Poole, J. Forrest Calland, Jianhong Lv, and Edward G. Chekan “Improving Surgical Pattern Recognition Through Repetitive Viewing of Video Clips” IEEE November 2004. [48] Kawai, T.; Westhofen, M.; Shibata, T “ Production and evaluation of stereoscopic video presentation in surgical training” Proceedings of the SPIE - The International Society for Optical Engineering, v 5291, n 1, 2004, 293-302
[49] www.downloads.com
[50] Seshadri S K, “Development of a virtual-reality based simulator for the da Vinci Surgical System”, Thesis, Feb 2008. [51] Seshadri S., Srimathveeravalli G., Kesavadas T., Guru K., “Design and Development of RoSS: A Training Simulator for daVinci”, Proceedings of Computer Assisted Radiology and Surgery (CARS), Spain, 2008. [52] Lendvay TS, Casale P, Sweet R, Peters C., “VR robotic surgery: randomized blinded study of the dV-Trainer robotic simulator”, Medicine Meets Virtual Reality 16, 2008. [53] G.S. Ruthenbeck, C.J. Carati, I.L. Gibbins, K.J. Reynolds, “A Virtual Reality 3D Jigsaw for Teaching Anatomy”, MMVR 2008.”, Medicine Meets Virtual Reality 16, 2008. [54] Pugh CM, Srivastava S, Shavelson R, Walker D, Cotner T, Scarloss B, Kuo M, Rawn C, Dev P, Krummel TH, Heinrichs LH., “The effect of simulator use on learning and selfassessment: the case of Stanford University's E-Pelvis simulator”, Medicine Meets Virtual Reality 11, 2001.
78
