Redesi gn of t he mot i on cont r ol l er gr i pper f or a vi t r eor et i nal sur gi cal r obot i c syst em
Daan Wilmink
BachelorAssignmentIndustrialDesign Engineering,August2013 University ofTwente
CONFIDENTIAL
Redesign of the motion controller gripper for a vitreoretinal surgical robotic system
D.G. Wilmink, s1010085
CONFIDENTIAL: do not make public before 12-08-2015
Industrial Design Engineering, Department of Engineering Technology
University of Twente
Examination date: 12-08-2013
Commissioned by:
PRECEYES Medical Robotics, Den Dolech 2 GEM-N 0.54 5612AZ Eindhoven
Examination committee:
dr. ir. W. Eggink
ing. T.G.M. Krone
ir. M.J. Beelen
Preface
The research described in this report is done in order to complete my bachelor of science in the eld of Industrial Design Engineering at the University of Twente. It contains a detailed process in the development of a technical product, enhancing various elds covered in the bachelor of this study.
For this research I would like to thank a couple of people. These people helped me during various phases within the research and were crucial in obtaining this result. First of all, I would like to thank ir. M.J.
Beelen for guiding me closely throughout the whole research and giving important and suggestions from PRECEYES Medical Robotics. It was great to work under your supervision at this innovative spin-o
at the Eindhoven University of Technology. You always took the time to answer my questions and gave useful tips and suggestions from an engineering point of view. Also, I would like to thank the other engineers at PRECEYES, dr. ir. G.J.L. Naus and dr. ir. H.C.M. Meenink, for their suggestions and tips during the research and design processes. It is great to experience a small glimpse of the passion that you put into the company.
During the analysis and design phase I had the support of several specialists. Dr. S. Keijser, thank you for letting me visit some of your operation procedures and taking the time for answering my questions.
Furthermore I would like to thank dr. E. van Oosterhout for taking the time to extensively test the dierent concepts and evaluating them from a clinical point of view, without you I would not be able to obtain a founded result. I want to say thanks as well to dr. ir. I.A.C. Soute for reviewing my testplan.
I learned a lot from you in the eld of the psychological/cognitive aspects of the design of technical products.
Last but not least, many thanks goes out to ing. T.G.M. Krone for guiding me through the research from the University of Twente and extensively reviewing the work I produced. Dr. ir. W. Eggink, thank you as well for taking the time to read my report.
Daan Wilmink
Summary
Vitreoretinal surgery describes surgery performed on the retina. A common procedure within this eld is vitrectomy, in which the liquid inside the eye is removed. Engineers at the PRECEYES Medical Robotics, a spin-o company at the Eindhoven University of Technology, developed a robotic system which assists the clinician in performing vitreoretinal operations. This robotic system contains a master controller and a slave (robotic arm). For the mastercontroller, the gripper requires a redesign that contains su cient inputs and enhances comfort and safety. The gripper should be equiped with an input for coupling the slave, actuation of the instrument and the movement of the center of motion of the slave in the xyz space.
A broad range of shapes are possible for the gripper, from ergonomical shapes following the curves of the hand to thin rotation symmetric shapes. During the design phase, several physical models are made and tested on a dummy of the master controller in order to nd possible comfortable shapes for the gripper. Four concepts substracted from these models are taken into an extensive user test with a vitreoretinal clinician. The concept that came out as a winner from this test, a gripper shaped as a bowling pin, is iterated and optimized to obtain a conclusive concept. Both the coupling and actuation button are placed on the bottom of the gripper so that they are controlled with the index and middle
nger respectively.
The detailing of the gripper involved the speci cation and mechanical integration of all the parts. A proportional acuation button is realized by a linear recessed button and a mechanical translation to a potentiometer. The coupling button exists of a physical module, shaped 180 degrees around the gripper, that transfers the forces exerted from each direction towards a push button through a liquid bag or gel.
Furthermore, a capacitive sensor is added at the thumb for safety so that the surgical system is only usable when the hand is correctly placed on the gripper. The parts and casing are produced out of aluminium since it is strong and corrosion resistant.
In the end a realistic design for a new gripper including required functions for new prototypes of the surgical system is obtained. However, some components like the capacitive sensor and coupling button needs further research and practical veri cation. When this is done, the design can be updated with new dimensions obtained from these components and transfered into a producable design including the tolerances and limitations of the machinery.
Samenvatting
Vitreoretinale chirurgie omvat de medische operaties die worden uitgevoerd op het netvlies in het oog.
Een veel voorkomende procedure hiervan is vitrectomie, waarbij de vloeistof in het oog verwijderd wordt.
Ingenieurs van PRECEYES Medical Robotics, een spin-o van de Technische Universiteit Eindhoven, hebben een robot ontwikkeld die de chirurg assisteert in het uitvoeren van vitreoretinale operaties.
Deze robot bevat een master controller en een slave (robotische arm). De mastercontroller heeft een herontwerp van het handvat nodig dat voldoende inputs heeft en de comfort en veiligheid in acht neemt.
Het handvat moet in het bezit zijn van een input voor het koppelen van de slave, het actueren van de instrument en het verplaatsen van het draaipunt van de slave in de xyz ruimte.
Een breed scala aan vormen zijn mogelijk voor het handvat, van ergonomische vormen die gevormd zijn naar de hand tot dunne rotatie-symmetrische vormen. Tijdens de ontwerpfase zijn verscheidene fysieke modellen gemaakt en getest op een dummy van de master om verschillende comfortabele vormen voor het handvat te achterhalen. Vier concepten die uit deze modellen naar voren zijn gekomen zijn uitvoerig onderzocht met een vitreoretinale chirurg tijdens een gebruikstest. Het concept dat hierbij als winnaar naar voren kwam, een handvat die gevormd is als een bowling kegel, is geitereerd en geoptimaliseerd om tot een eindconcept te komen. Zowel de koppel- als de actuatieknop zijn onderaan het handvat geplaatst zodat deze worden aangestuurd door respectievelijk de wijs- en middelvinger.
Het detailleren van het handvat omvatte de speci catie en mechanische integratie van alle onderde- len. Een proportionele actuatieknop is gerealiseerd door een lineaire verzonken knop en een mecha- nische translatie naar een potentiometer. De koppelknop bestaat uit een fysieke module, 180 graden gevormd om het handvat, die uitwendige krachten van alle richtingen geleidt naar een drukknop door een vloeistofzakje of een gel. Daarnaast is er een nabijheidssensor toegevoegd voor de veiligheid bij de duim zodat de robot alleen te gebruiken is wanneer de hand op een correcte manier op het handvat geplaatst is. De onderdelen en behuizing wordt gemaakt uit aluminium omdat dit een sterk en roestvrij materiaal is.
Uiteindelijk is er een realistisch ontwerp voor een nieuw handvat inclusief de benodigde functies naar voren gekomen. Echter, sommige onderdelen zoals de nabijheidssensor en de koppelknop moeten verder onderzocht en getest worden. Wanneer dit is gebeurd kan het ontwerp gecomplementeerd worden met exacte afmetingen van deze onderdelen en kan het omgezet worden tot een produceerbaar ontwerp inclusief de toleranties en limitaties van de machines.
CONTENTS CONTENTS
Contents
Preface 4
Summary 5
Samenvatting 6
1 Introduction 10
2 Analysis 11
2.1 Vitreoretinal surgery . . . 11
2.2 PRECEYES Surgical System . . . 12
2.2.1 Technical speci cation of the gripper . . . 13
2.3 Other surgical robotic systems . . . 15
2.3.1 Da Vinci Robotic System . . . 15
2.3.2 Robotic Systems at St Hopkins University . . . 16
2.3.3 Laser robotic system at CMI Montreal . . . 17
2.3.4 Conclusions . . . 17
2.4 Other vitreoretinal robotic systems . . . 18
2.4.1 University of Tokyo . . . 18
2.4.2 Johns Hopkins University . . . 19
2.4.3 Other systems . . . 19
2.4.4 Conclusions . . . 20
2.5 Other precise professions and products . . . 20
2.5.1 Tattooists . . . 20
2.5.2 Artists . . . 21
2.5.3 Products with similar grip . . . 21
2.5.4 Conclusions . . . 22
2.6 Operation procedure . . . 22
2.7 Current instruments . . . 25
2.8 Technical functions of the gripper . . . 26
2.8.1 Input possibilities . . . 27
CONTENTS CONTENTS
3 Concept generation 30
3.1 Idea generation . . . 30
3.1.1 Sketching . . . 30
3.1.2 Physical ideas . . . 34
3.2 Concept generation . . . 37
3.2.1 Concept 1: Shaped-to-the-hand . . . 37
3.2.2 Concept 2: Cone shaped . . . 38
3.2.3 Concept 3: 'Hanging' hand . . . 39
3.2.4 Concept 4: Thin shaped . . . 40
3.3 Testing and evaluation . . . 41
3.3.1 Theoretical Preparation . . . 41
3.3.2 Execution . . . 42
3.3.3 Results & Conclusions . . . 44
3.4 Iteration of the design . . . 46
3.5 Conclusions and choice of concept . . . 47
4 Detailing 49 4.1 Integration and speci cation of the buttons . . . 49
4.1.1 De ning the actuation button . . . 49
4.1.2 De ning the coupling button . . . 51
4.1.3 Exact shape and placement of both buttons . . . 52
4.2 Integrating the safety aspects . . . 54
4.3 De ning the core material . . . 55
4.4 Cognitive usage . . . 56
4.5 Manufacturing . . . 56
4.6 Construction and detailed representation . . . 57
4.6.1 Mechanical Integration . . . 58
4.6.2 Parts overview . . . 60
4.6.3 Interface with the master . . . 61
4.7 Additional input: coupling xyz . . . 61
4.8 Conclusions . . . 62
CONTENTS CONTENTS
5 Conclusions and Recommendations 63
5.1 Conclusions . . . 63 5.2 Recommendations . . . 63
References 65
Nomenclature 67
Appendix A 68
Appendix B 76
Appendix C 78
Appendix D 80
Appendix E 84
Appendix F 90
Appendix G 92
1 INTRODUCTION
1 Introduction
In the medical world, often di cult surgical treatments have to be performed. Vitreoretinal surgery, which means surgical operations performed on the retina inside the eye, may be one of the most precise medical elds. As a consequence, students at the Eindhoven University of Technology designed and developed a robotic system that assists the clinician in performing these operations in a more precise way. This resulted in the spino company called 'PRECEYES Medical Robotics' that tries to further improve this robotic system and commercially distribute it in the medical world.
However, some parts of the robotic system are still under developed and need further improvement before the device can be distributed commercially. One of these parts is the so-called gripper, which is part of the surgeon's console, the motion controller, that is held by the user to provide position commands to the instrument manipulator. Therefore the assignment was given to redesign this gripper, given the prospected target group and speci cations. This resulted in a complete development and design process that has lead to a realistic concept.
The goal of this report is to present and substantiate the development process of a gripper for a vitreoretinal surgical robotic system. The rst part of this research consists of a broad analysis exploring this particular eld of surgery and their clinicians. Furthermore, a study is done on the PRECEYES Surgical System as on other systems and solutions for similar procedures. This analysis leads to concrete requirements that de ne the barriers of the design. Then, an iterative process starts with idea generation and physical modelling to generate a number of suitable dierent types of concepts. Veri cation and selection of these concepts is done through several user tests. At the end, the developed concept will be converted into a realistic design that can be used as a framework for the production of a working prototype.
The complete process will be divided into three parts; analysis, concept generation and detailing. Each part will be concluded with a brief overview of the results and an evaluation of these results.
2 ANALYSIS
2 Analysis
This rst section is very important in the design proces. A closer look is taken on both the target group as the company for which the assignment is done. Also, similar areas of interest are studied in order to get a proper overview of all the possibilities and suitable solutions. The main goal of this analysis is to collect concrete information that can be used in the creation of a list of requirements. This list of requirements will be the foundation of the subsequent concept generating phase. First, a study will be done about the eld of vitreoretinal surgery and the robotic system developed by PRECEYES Medical Robotics.
2.1 Vitreoretinal surgery
In order to get a complete understanding of the surgical robot and the needs of the users, it is necessary to get a brief overview of vitreoretinal surgery. Therefore, a small study has been made about this medical eld.
Vitreoretinal surgery is the term that is used to indicate medical treatments on the retina and liquid (vitreous) inside the eye. Since the retina, a light-sensitive layer of tissue, is located in the inner surface of the eye, this treatment has to be carried out inside the eye of the patient. There are several retinal diseases one can be treated for; the most prominent of these are: macular degeneration, retinal detachment, macular holes, and diabetic retinopathy [2]. Such treatments are often carried out by ophthalmologists, specialists in medical and surgical eye problems, that are focussed on vitreoretinal surgery. This degree has to be obtained in a fellowship as a follow-up of the MD-degree [3]. Therefore, the total education track can take more than ten years and thus makes this profession a true specialism.
Figure 1: Representation of a vitreoretinal procedure. An instrument (upper gauge) and light source (lower gauge) is inserted into the eye.[1]
The rst successful case of vitreoretinal surgery came from the Swiss ophthalmologist Jules Gonin in 1925 [4]. Ever since, vitreoretinal surgery process has been optimized. By inserting two gauges in the eyeball,
2.2 PRECEYES Surgical System 2 ANALYSIS
use of any sutures afterwards. Besides, a cannula is placed in the incision which allows the specialist to enter the eye with dierent instruments via the same incision. This system, which reduces as well as surgical trauma as operating time, is still used today by many clinicians.
In paragraph 2.6 a more detailed description will be made about the current operation procedures in the
eld of vitreoretinal surgery.
2.2 PRECEYES Surgical System
The spin-o company PRECEYES developed a robot with the purpose to execute vitreoretinal surgery.
The robot makes use of a master-slave system. This means that the movement of the master device, the sugeon's console, will be translated to movements in the surgical robot, the so-called `slave'. With this system, properties like scaling movement, ltering vibrations of the hand and feedback (e.g. about the force one exert) are enabled that enhance medical precision
Figure 2: Usage of the PRECEYES Surgical System [5]
Figure 3 shows the slave part of the robot. This part consists of a robotic arm that can rotate around the three axes Ψ (psi), Φ (phi) and Θ (theta), respectively the x, y and z-axis. The instrument, attached at the end of the arm, can move in the z-direction as well, giving the slave a 4 DoF (degree of freedom) in total. The three axes Ψ, Φ and Θ are coinciding each other in one point (see gure 3). This intersection creates a xed center of motion and thus an ideal point suited for placement at the cannula is obtained.
2.2 PRECEYES Surgical System 2 ANALYSIS
Figure 3: Representation of the slave (left) and the master (right)
The master part of the current robot is shown in gure 3 as well. This is the part that functions as the motion controller for the slave. This system has various axes as well (ψ, φ and θ, now noted in small letters) obtaining a similar intersection as with the slave. This intersection also forms a center of motion (remote center of motion, RCM). Since the arm of the controller is now attached in the opposite way, the user controls the slave as one controls the tip of the surgical instrument in the patient's eye. The gripper is attached at the end of this arm and is provided with a button. When holding this button, the slave can be controlled. That is, motions around the cannula can be made as well as the movement of the instrument in the z-direction. One argument to hold the button before moving the slave came as a consequence of the scaling -compare it with a computer mouse that has to be lifted when moving in large distances on the screen- as well as the safety of the patient. This way of movement will be refered to as 'clutching'.
Since this robot can assist the clincian in operating ten times more precise via scaling and ltering [6], it can be of great use in vitreoretinal surgery. In fact, PRECEYES designed the robot to enable new procedures in this eld of surgery. For instance, treatments in the veins of the retina (e.g. removal of a blood clot) cannot be executed by hand since it requires an excessive amount of precision. Since the gripper is the core part for this assignment, the next sub-paragraph will explain some detailed speci cations about this part of the robotic system.
2.2.1 Technical speci cation of the gripper
The current design of the gripper, as is shown in gure 4, is made very robust and is also ergonomically not studied and complete. The robust shape is a consequence of the motor that is integrated in the gripper. This motor is used to give feedback to the trigger so the clinician knows how much force he/she
2.2 PRECEYES Surgical System 2 ANALYSIS
Figure 4: External design and usage of the current gripper (left) and a section view of the current gripper (right) [8]
The motion freedom of the gripper, the free space in which one can use and rotate the gripper over all the axes, is very important for the physical design as well. Namely, in the complete space in which the gripper is used, an equal ergonomically condition is desired for the user. The speci cations of this motion freedom can be seen in table 1.
Axis Current limitation) Desired limitation
Φ φ m1a 175, 8◦ 50◦
Ψ ψ m1b 70◦ 50◦
Z z m1c 16mm 50mm
Θ θ m1d 360◦ 140◦
Table 1: Overview of the current limitations considering the motion freedom [7] and the desired limita- tions
2.3 Other surgical robotic systems 2 ANALYSIS
Figure 5: Nominal position of the gripper (left), visualized area of the motion freedom (right).
The values shown in table 1 represent the total motion freedom around the nominal position. The position in which this is the case can be seen in gure 5. The desired limitation represents the motion freedom that is considered more convenient. The current limitation is enabled mechanically, whereas the desired limitation is also enabled currently through software-based 'airbags'. Airbags means limitation performed by the electrical motors exerting force in the opposite direction. A 3D-view on the motion freedom can be found in gure 5 as well. This visualization will be of use in the evaluation of the dierent concepts.
2.3 Other surgical robotic systems
In order to gain a better insight into dierent solutions that can be made for the gripper of a medical robot, a competitive study is done. This study can give insight in what solutions are already on the market, what solutions are working and what solutions are protected through IP's. First, some prominent robotic systems in the medical eld will be discussed. Then, a study is done on other robotic systems that are being developed in the eld of vitreoretinal surgery.
2.3.1 Da Vinci Robotic System
The da Vinci robotic system may be the most known example of a medical robotic system. It is used to perform the minimally invasive procedure, which means that it is less invasive than traditional surgery
2.3 Other surgical robotic systems 2 ANALYSIS
Figure 6: Overview of the Da Vinci Robotic System (left) [10] and the controller (right) [11]
The controller, for this assignment the most relevant part, consists of a console that is supplied separately from the main robotic system. The instrument is controlled by placing the thumb and pointing nger into simple bands, as one can see in gure 6. Pinching the ngers is used for actuating the instrument (e.g.
grabbing tissue). Global movement of the instrument can be enabled by making the same movement with your hand; arms and joints rotating in seven degrees of freedom will detect the performed movement and translates it to the robotic system enhancing great precision and exibility. Additionally, the ellbows of the user are supported by a beam attached to the console. A downside of this system is that it is patented and therefore hard to implement in an own commercial product.
2.3.2 Robotic Systems at St Hopkins University
Another example of a medical robotic system is the system developed at the St Hopkins University.
This robotic system is developed in order to operate more eectively in a patient's throat. It consists of a snake-like robot that can move through small funnels (e.g. the patient's throat). However, the control method is roughly the same as it is with the da Vinci robotic system using a pinching motion for actuating the instrument and several robotic arms that registrate the spatial movement of the hand.
Since most emphasis is put onto the slave-part of the robotic system, most research facilities uses standardized masters to control their robots.
2.3 Other surgical robotic systems 2 ANALYSIS
Figure 7: Controlling a robotic system developed at the Johns Hopkins University [13]
2.3.3 Laser robotic system at CMI Montreal
A third example of a medical robotic system has been found at the Computer Medical Institute of Montreal. This system has originally been developed with the purpose of more eective laser phonomi- crosurgery. The controller, however, consisted of a simple joystick which caused big problems when large precision movements were made. Therefore, the institute developed custom software for the joystick which also incorporated a playback function. This function enabled the clinician to memorize particular target points and reposition the laser to these exact points [12].
Figure 8: Handling of a laser robotic system (left) and the interface of the developed software for this system (right)
2.3.4 Conclusions
2.4 Other vitreoretinal robotic systems 2 ANALYSIS
shows the big advantage proper software can make for a robotic system. Similar functions and software could be developed for the PRECEYES system as well, but it is not within the assignment to do further study in this area.
2.4 Other vitreoretinal robotic systems
Now a closer look is made on similar research inititatives as the PRECEYES Surgical System. What other developments are currently being made in the eld of robotic vitreoretinal surgery? But most of all, in what way do they integrate the surgeon's console and interface? First of all, it is important to point out that common robotic systems can be divided into three categories: co-manipulation devices, handheld tools and master-slave systems. The rst describes the case in which the manually used instrument is co- used by a robot that therefore enables automation, force feedback and most importantly steadiness. The second indicates handheld systems in which the robotic part is integrated and therefore steadiness can be enhanced. The third and latter system focuses on systems where the surgeon controls a robot/controller that translates the motion to a second robot. As mentioned earlier, this principle is also used by the PRECEYES system.
2.4.1 University of Tokyo
At the University of Tokyo, another master-slave vitreoretinal surgical robotic system is developed. This slave of this system is quite similar to the PRECEYES system. However, the controller is using a dierent approach. As one can see in gure 9, this controller looks quite similar as the one used in the da Vinci Robotic System. Here, the user controls the slave separately (i.e. dierent location) by looking on a digital screen. The motion of the hand is traced by a robotic arm that consists of several sub-sections that is therefore able to register the complete movement that is made in the free space.
Figure 9: Vitreoretinal surgical robotic system at the University of Tokyo [14]
An interesting research that is done by this university can be found in [14]. It shows a large gain in precision when using the robotic system. Besides, the dierence of precision between the clinician and
2.4 Other vitreoretinal robotic systems 2 ANALYSIS
the engineering students is less. This could mean that the learning curve for performing professionally through such a system is less steep than that of a manual procedure.
2.4.2 Johns Hopkins University
Where in the previous paragraph already an example of a medical robotic system at this university was discussed, the Johns Hopkins University also developed a system for vitreoretinal surgery. This robotic system consists of the ordinary instrument combined with a robot which lters frequencies in order to enhance a steady hand. Due to the construction of this co-manipulation device, the controller is made similar to the original instrument. In other words, it is shaped as a small cylinder since it uses an original instrument which is connected to a robot.
Figure 10: Steady hand robotic system at the Johns Hopkins University [15]
2.4.3 Other systems
There are more vitreoretinal robotic systems that could be discussed. However, these are all in an earlier stage of development. Developers and researchers seem to focus more on the slave system, the executing part, of the robot. The system is then controlled from prefab masters/controllers. In gure 11 still some examples of other vitreoretinal robotic systems can be found.
2.5 Other precise professions and products 2 ANALYSIS
Figure 11: Concepts of vitreoretinal robotic systems at the Jules Stein Eye Institute [17] and the Vanderbilt University [16]
2.4.4 Conclusions
Some interesting conclusions can be drawn from this study. First of all, one can see that the PRECEYES system is relatively far developed in this eld. The biggest competitors can be found at the Tokyo University and the Johns Hopkins University. The Tokyo University uses a controller which is quite similar to the da Vinci Robotic System and thus a fairly proper way of controlling, but the construction is simply too dierent from the current construction of the master within the PRECEYES Surgical System. While the robotic system of the Johns Hopkins University uses a dierent approach, the design of the gripper can be of interest for the PRECEYES Surgical System. It shows that a robotic system can be perfectly combined with the ergonomics of the original tools. However, one must be careful not to forget that this system does not integrate scaling, while the PRECEYES system does.
2.5 Other precise professions and products
Not only in the medical eld, but also outside this eld professions that require similar precision can be found. The aim in this study is to nd solutions in similar products in other professions. Questions like
In what professions a similar tool is used in order to enhance similar precision? and How does the controller/gripper of these products look like? are tried to be answered.
2.5.1 Tattooists
A tattooist is a nice example of a profession in which precise work has to be executed via a semi-robotic system. The physical design of the `controller' is not very impressive or ergonomically, and the injector of the machine can be controlled via a foot switch. However, it seems like precise work indeed can be executed with a simple physical design which, besides, suits a broad type of user (e.g. left/right-handed).
2.5 Other precise professions and products 2 ANALYSIS
Figure 12: Design of a tattoo machine (right) and the hold onto the machine (left).[9]
2.5.2 Artists
An obvious example of precise work is art. An artist must possess a very steady hand, and thus requires some similar skills as the clinicians. Because of the incredible amount of people that produces artwork on regular basis, an immense amount of specialized products and pencils are available. A simple search on the internet can provide concrete ergonomic pencils designed in special to prevent fatigue.
2.5.3 Products with similar grip
In the previous paragraph, the pencil was already pointed out as a product of which the ergonomic design can be of great bene t during the design of a new gripper. A graphical representation can be found in gure 13 in order to get an overview of these similar products. Here, one can observe that a lot of these products try to unstress the hand. For instance, cutouts as in the pencils on the bottom right decrease the intensity of pressure points (e.g. a point of contact with the nger is converted into an bigger area of contact). Structure of the gripper's material, is with the gripper on the bottom left in
gure 13, can increase transverse friction so the hand maintains its position better.
2.6 Operation procedure 2 ANALYSIS
Figure 13: Overview of several grippers designed especially to prevent fatigue.
2.5.4 Conclusions
What concrete aspects can be learned from other similar professions and tools? Products shaped for the hand as is shown in gure 13 can reduce the risk of fatigue and cramp, but if it really enhances greater precision is not a guarantee. If this was the case, tattoo artists and professional artists may use curvy shaped pencils/grippers instead of simple, small and straight shaped tools more. However, shaping the gripper ergonomically to the hand is a cognitive bene t. The products in the listed gure shows an obvious way of placing your hand on the gripper; the shape forces the user to handle the gripper correctly. The downside is that it has a negative eect on the exibility. In the case of the gripper of the robotic system it has a negative eect on, for instance, the rotation of the gripper between the ngers.
It is important that as well as ergonomically shaped-to-the-hand grippers as simple straight grippers are developed in the concept phase and tested in a proper experiment involving great precision. Subsequent, a relative importance can be graded on which the gripper can be tested during the evaluation.
2.6 Operation procedure
In order to gain better insight in the user, the clinician, three operation procedures were visited in the academic hospital in Nijmegen, the Netherlands. This paragraph will give a brief overview of the conclusions and observations that were made that day.
The procedures that were executed were vitrectomies. In this procedure, the liquid inside the eye is removed in order to, in this case, reattach the retina. The operation procedure is usually executed by the clinician and is assisted by one or two nurses (depends on which eye is treated) that are handing over instruments and moistening the eye. Namely the instruments are located on at a xed side of the operating table and thus an extra nurse for handing over the instruments is needed when the eye at the
2.6 Operation procedure 2 ANALYSIS
opposite side is treated. The clinician sits at the head-end of the operating table and the assisting nurse sits on the side of the eye that is treated (see gure 14). Most of the vision is done through a microscope that is mounted above the patient's head. This microscope is handled by the clinician through multiple foot pedals. The clinician sits in an ergonomically proper position with a straight back, correct height of the seat and a correct height of the microscope. Consequently, the lower arm makes a 90 degrees angle with the upper body and thus enhances optimal control (see gure 15).
During the whole procedure, a series of tools is used. At rst, the clinician needs to anesthetize the eye.
This is done by using a tweezer to tilt the top of the tissue a bit so subsequently an injection can be done in this tissue. The cannula is placed using a similar tool as in gure xc. Next, the liquid inside the eye is removed using, in the one hand, a light source (shaped similar as in attachment B) and, in the other hand, a small pump; the clinician at the hospital used the tool in gure xa. When the liquid is removed, one can start searching for the detachment and subsequently laser the desired area. This is done as well with a light source, combined with a laser tool in the other hand which has a similar shape. Having completed the lasering process, the eye can be lled with a gas. This is done with a similar shaped tool as well. Afterwards, all the tools and cannulas are removed and the eye is covered up with bandage.
Figure 14: Top view of the setting during an operation procedure. Green is the clinician, red is the patient and blue is the assisting nurse.
2.6 Operation procedure 2 ANALYSIS
Figure 15: Side view of the setting during an operation procedure.
Such an operation procedure takes approximately 45 minutes on average. During the procedure, the hands of the clinician are somehow supported by the patient's head. The position of the hands in relation to the head can be adjusted carefully by using the little ngers. However, this could be a personal preference. The overall movement is performed by holding the instrument with all ngers except the little nger and mostly using the lower arm to enable rotation. Additionally, the clinician remarked that sometimes he uses an extra nger (middle nger) to move the eye a little bit and that the tools can be used to move the eye around. In this way, an dierent degree of insertion is made and other corners in the eye can be reached. Also, he preferred to use thinner instruments to avoid pressing in the instrument too hard. Probably this is because of the habit of using thin instruments and the larger area the feedback of the nger has to travel through. Another important fact is that some instruments have an asymmetric end (e.g. a curved tube) or an opening for injection at one side so that rolling the instruments between the ngers may be necessary.
2.7 Current instruments 2 ANALYSIS
Figure 16: Real-life image of a clinician performing vitreoretinal surgery.[19]
2.7 Current instruments
When a study is made in the current vitreoretinal product market, one can distinguish several players in this market. Alcon possesses a big share in this market, producing several complex systems that are used in standard procedures. Another company that produces these instruments is the Dutch Ophthalmic Research Center (DORC). They produce separate handheld instruments for vitreoretinal surgery.
Dierent types of instruments are used throughout the operation procedures. Importantly, these dierent instruments are accompanied with dierent shapes. In order to give a brief overview of the used types of instruments during the operation procedure, an overview can be found in attachment B. Here, one can see that the vitrector is the only instrument that has a thicker shape than the other instruments (Ø>10mm). The questioned clinician indicated that he disliked the thicker instrument since it causes him to pinch harder in the gripper and therefore increases the risk of fatigue and decreases the amount of precision. This could be explained by the fact that the clinicians uses the thicker instrument for only one period (approximately 15 minutes) during the whole operation (40-60 minutes) by one hand. Thus around 15% of the time a thick instrument is used, whereas 85% of the time a thin instrument is used.
It could thus be stated that clinicians are used to thin instruments. In gure 17, the usage of certain instruments during the operation can be viewed.
2.8 Technical functions of the gripper 2 ANALYSIS
Figure 17: Position of the clinician and the grip of the hands during a vitreoretinal operation procedure.[20][21]
2.8 Technical functions of the gripper
Not only a physical design of the gripper has to be made. Also, the gripper has to be integrated properly onto the master device. Therefore, an analysis of the technical requirements has to be done in order to avoid any misalignment from occurring during the integration and to assure that every function is taken into account during the development. First of all, there are several functions the robot has to execute or may execute in the future. These functions are listed below:
• The slave can rotate around the cannula in all DoF.
• An extra function of the instrument can be activated (e.g. pinching a tweezer).
• The slave gives feedback to the master when reaching the edge of the motion freedom. This is currently done using 'airbags', as discussed in subparagraph 2.2.1.
• The slave gives feedback of the pressure force to the master controller. That is, when moving the instrument along the z-axis and the tip is exerting a force on tissue, this force is translated back to the user.
• Center of motion of the slave can be moved to dierent locations in xyz-space. Purpose of this is to quickly move the eye. This functionality and purpose is similar to the current manual method of rotating the eye discussed in paragraph 2.6. Currently moving the RCM can be done only manually. However, a system for automatic movement within this area is under development.
The feedback of the force exerted upon the tissue is now communicated via a motor connected to the button, which explains the thickness of the current controller. However, this feedback could be communicated dierently as well, so that the size of the controller can be reduced. Therefore, it is important to keep these functions abstract and not to restrain the physical design too much.
2.8 Technical functions of the gripper 2 ANALYSIS
2.8.1 Input possibilities
Since there are many ways to create an input besides plain buttons, it is useful to create a brief overview of these possibilities. When a proper overview is created, substantiated decisions can be made in this area. First, the dierent types of inputs are de ned:
• Coupling: equal to the input on the current gripper. Upon activation, the master couples to the slave and translation of the movement is enabled.
• Actuation: here, a function of the instrument is activated (e.g. pinching forceps).
• Coupling xyz: as discussed in the introduction of this paragraph, this system is under development and therefore may need input in the future.
The most important function that the gripper has to execute is the coupling of the slave. Since this input must be activated during relatively long periods, no fatigue must occur during the use. During the operation procedures it is observed that the thumb and the index nger are the only ngers that are continuously on the instrument. This counts to a certain degree for the middle nger as well, but for some short moments this nger was detached from the instrument. Furthermore, these ngers are also relatively from each other placed on the same position. It thus can be said that this function must be executed with the a) thumb, b) pointing nger or c) both. For each possibility the following potential inputs can distinguished:
a) Button or sensor b) Button or sensor
c) Pinching, buttons and/or sensor(s)
However, it must be remarked that if both ngers are used for this input, the only way to still use (one of) these ngers for other inputs is by using only the lower parts of the ngers (as is possible with pinching).
2.8.2 Safety aspects
An important aspect that goes along with these functions is the safety. When designing a new controller for a surgical robotic sytem, safety has to be taken into account. To get a proper overview of all the incidents that could occur with the controller, and consequently substract useful information and concrete solutions, a failure mode and eect analysis (FMEA) has been done on the controller. Table 2 shows a summary of the FMEA, the details of this analysis can be found in attachment C.
2.9 Conclusions and requirements 2 ANALYSIS
Failure Mode RPN Action (Yes/No)
Clinician bumps against the controller 126 YES
Slave gets activated by accident 189 YES
Instrument makes unmeant movements 48 NO
Slave changes position without meaning to 56 YES Old instrument is still attached to the slave 24 NO Instrument drains liquid at the wrong side 42 NO Gripper detaches from the master device 14 NO
Gripper chrushes/breaks 16 NO
Table 2: Summary of the FMEA
In the FMEA, dierent ways in which the gripper could fail and cause serious consequences are described.
Furthermore, a value is given to important aspects like 'severity', 'occurance' and 'detection'. These values create a risk priority number (RPN) that indicates the priority of the given mode of failure. The last collumn indicates wether or not action will be undertaken in the design proces. However, since in this case a comparison is done with the manual way of operating, some aspects do not necesarily needs action but needs to be translated into the new gripper design of the robotic system. For instance, the fact that the physical shape diers for each instrument for the current situation does not yield for the gripper on the robotic system; the gripper stays the same for each instrument. These are aspects that, despite the FMEA says 'no action', still needs to be taken into account.
When 'YES' is submitted into the last collumn, de nite measures needs to be undertaken. These are mostly the failures with the highest RPN. For instance, 'activating the slave by accident' can indicate that a proper study needs to be done on the button/switch that will be used and the required force to activate such a button. Besides, several of these failures indicate that multiple button/sensors may be needed in order to activate the slave and still enhance full safety.
2.9 Conclusions and requirements
Before the de nite list of requirements can be set up, brief conclusions of the analysis needs to be clari ed. First of all, we can say that there is not much development in the eld of robotic vitreoretinal surgery so there is a large likelihood that the end design is a true unique one in its solutions and physical design. Next, it was seen that similar tools for other purposes can be of great contribution for the physical design. However, requirements of the clinicians and the company must be taken into account.
For the clinicians, it is important that the gripper not only enables precise work but also enhances freedom in handling the gripper. For example, the questioned clinician found thinner instruments more convenient to use. Another important aspect is that all the functions can be used properly and convenient. Since the clinician only uses one gripper instead of multiple tools, there is a loss of direct and cognitive natural handling. For instance, the usage of a forceps is obvious and natural (i.e. direct mechanical translation of the pinching movement) but completely dierent than the usage of a vitrectrome. However, now both instruments are controlled by the same gripper, and thus usage of the may become more di cult.
For PRECEYES Medical Robotics, on the other hand, it is important that the gripper communicates properly with the slave part of the robotic system. A series of requirements can be obtained from the
2.9 Conclusions and requirements 2 ANALYSIS
current robotic system, and the inputs it requires. The robotic system will be a new way of performing this kind of surgery. For instance, normally (manual method) the RCM is located below the hand at the end of the gripper, whereas at the robotic system the RCM is located above the hand. In any way, the handling of the system diers from the manual method so clinicians will have a learning curve in handling the robotic system in an optimal way.
2.9.1 Requirements
1. The gripper can enable a 20 degree rotation around the theta-axis without detaching the hand from the gripper.
2. Through clutching, the gripper can enable a 180 degree rotation around the theta-axis without detaching the hand from the gripper.
3. The gripper must enable an equal ergonomically condition for both the left hand and the right hand.
4. The gripper is able to couple the master with the slave.
5. The gripper is able to couple the master with the slave for the movement of the RCM in the xyz-space.
6. The gripper is able to activate an extra function (actuation).
7. The actuation input is proportional, so that a variable output is possible.
8. The gripper can still be used when gloves and a sterilization bag are used.
9. No fatigue occurs when operating the gripper for 45 minutes.
10. The precision of the robotic system remains the same or increases with the new gripper opposed to the system with the old gripper.
11. The little nger is not required to be on the gripper (has no essential function).
12. The functions are integrated cognitively (3 out of 4 persons guesses/uses the functions right the
rst time).
13. There are no sharp edges or corners on the gripper.
14. The gripper enhances an equal ergonomically condition throughout the whole workspace.
15. The gripper is resistant against pressure forces up to 68kg.
16. The attachment of the gripper to the master device is resistant to stretch forces up to 1000N.
3 CONCEPT GENERATION
3 Concept generation
In order to generate proper concepts, seperate phases must be distinguished in this section. First, general ideas and designs are set on the paper in order to nd possible solutions for the gripper. When this is done, an overview is created of the directions one can go to with the design. In this way several distinguished concepts can be created that covers seperate elds of solutions.
3.1 Idea generation
The generation of ideas for the gripper is mainly focussed on the physical shape. These ideas are very broad and are used to create an overview of possible shapes. This does not mean that detailling like technical speci cations are not taken into account at all, but the emphasis on this area lies in the stage were the physical shape is more de ned. The process towards the concepts is chronologically as follows:
1. Sketching: basic ideas are set on paper.
2. Converging of sketches: promising sketches are further explored and developed.
3. Physical Ideas: 'shaping', ideas are created by physical modelling and modifying material.
4. Iteration: results of point 2. and 3. are all produced into physical models. Then, these models are tested, modi ed, tested, modi ed etc. untill a satis ed optimal model is obtained.
3.1.1 Sketching
A quick and easy way to get a proper overview of dierent solutions and shapes is through sketching.
First, a series of sketches are made to analyse dierent types of shapes and instruments. Interesting shapes are marked and taken to the next stage for elaboration. Note that the shapes of these sketches are drawn as instruments, with a gauge connected to the end. This is done purely to show the front- end and enhance better visualization, the actual concept will not contain a gauge (purely the gripper).
Consequently, in these sketches the center of motion will not be at the beginning of the gauge (as is the case with manual surgery) but on the back (right side) of the instrument.
3.1 Idea generation 3 CONCEPT GENERATION
3.1 Idea generation 3 CONCEPT GENERATION
Several drawings have been made about the position of the hand with regard to the gripper ( gure 19).
The position of both the hand with manual instruments and the hand on the robotic system is drawn.
The dierence is that in the rst case the instrument lies against the joint of the thumb and the pointing
nger, and in the latter case the 'instrument' (gripper) is attached to an external point.
Figure 19: Position of hands during the handling of dierent types of instruments/systems. Topleft and topright: handling of conventional manual instruments. Bottomright: handling of the motion controller.
Bottomleft: handling of an ergonomically shaped pencil.
Promising shapes that came out of the sketches are taken into further development. Small iterations were made in order to explore this particular shape further. Then, other directions for a particular characteristic of that shape are explored (e.g. front-section-view in order to explore round or cubical shapes). At the end, a visualisation is drawn of the usage of that chosen shape. Note, here the same sketch as in gure 19 is used.
3.1 Idea generation 3 CONCEPT GENERATION
3.1 Idea generation 3 CONCEPT GENERATION
Figure 20: Detailing of promising shapes obtained from the sketches.
3.1.2 Physical ideas
Sketching is not the only method that is used to create suitable concepts. Since the shape of a gripper is hard to de ne and quantify, an iterative physical study is done to get the 'feel' of these shapes. To simulate an equal condition as with the original robotic system, and since the original robotic system is too expensive to modify, a 'dummy' is made by using an older prototype of the master part of the robotic system. On this dummy, several physical models can be attached and immediately tested in a similar condition as the real one.
3.1 Idea generation 3 CONCEPT GENERATION
Figure 21: The dummy created for testing the physical concepts
The creation of these physical ideas consists of several steps. At rst, a small hole is drilled into a simple rectangular cube so it can be connected onto the dummy. Then, the iterative process starts by
uctuating between testing and polishing/crafting the gripper untill a suitable 'idea' is created.
3.1 Idea generation 3 CONCEPT GENERATION
Figure 22: Usage and testing of produced models. Testing the models in the workshop enables an iterative development.
Additionaly, promising shapes that were found in the sketches are made into a physical model as well.
Together with the physical ideas they form a broad collection of concrete potential concepts. Among these, the models that were experienced as the most convenient during test sessions on the dummy and furthermore belongs to each an own segment of ideas (e.g. shaped-to-the-hand, out-of-the-box) were taken to the next stage; concept generation.
Figure 23: Physical models of the gripper created during both the sketching as the physical idea gener- ation phase.
3.2 Concept generation 3 CONCEPT GENERATION
3.2 Concept generation
Before the concepts can be compared and a de nite choice can be made, the chosen ideas must be developed to detailed concepts. In this case, concrete placement of the hand as well as the placement and functions of the buttons is of great concern. Results and comparable drawings can be found in subparagraph 3.2.1 to 3.2.4.
3.2.1 Concept 1: Shaped-to-the-hand
Figure 24: Visualization, details and dimensions of concept 1.
This concept came out of the physical ideas phase. It is made in such a way that it ts properly to the
ngers that are attached to the gripper throughout almost the entire motion freedom. The tting is achieved by notches made onto the surface of the model; one for the thumb two for the pointing and middle nger. Since it is shaped to the hand, an other version will be needed when the other controller is used with the left hand.
+ Ergonomic placement of the ngers
3.2 Concept generation 3 CONCEPT GENERATION
3.2.2 Concept 2: Cone shaped
Figure 25: Visualization, details and dimensions of concept 2.
This shape already came forward in the sketching phase. However, after physical modelling the shape it was found out that one always tends to place their ngers on the thinner part of the model. Therefore, the placement of the button has changed. Another interesting feature is the rotating end of the model.
This feature enables the user to rotate the instrument around the z-axis without turning the button-part (and thus enhances continuous actuation). This feature is not necessarily binded to this concept, it can also be applied to other concepts. If this feature is really a proper and progressive solution will be considered at the testing and evaluation phase.
+ Fingers can be placed on preferred manner/thickness + Smooth surface, so rolling between ngers may be easy + Provides enough space for functions
- Placement of buttons must be well considered
3.2 Concept generation 3 CONCEPT GENERATION
3.2.3 Concept 3: 'Hanging' hand
Figure 26: Visualization, details and dimensions of concept 3.
This is an out-of-the-box concept which is totally dierent than the other concepts, but can lead to interesting results. It is inspired on an existing pencil design that claims to have an ergonomically superior design. Since the nger is 'stuck' in the model, rotation around the z-axis can be more di cult in the classical manual way. Therefore, the same principle as the previous concept is used to rotate the instrument.
+ Ergonomic natural position of the hand - Placement of buttons fairly limited - Turning around z-axis somehow limited
3.2 Concept generation 3 CONCEPT GENERATION
3.2.4 Concept 4: Thin shaped
Figure 27: Visualization, details and dimensions of concept 4.
In order to obtain a concept that is closely related to current manual instruments, this concept is developed. The shape shows similarities and enhances the same grip clinicians are currently used to.
However, wheter or not this grip is really as convenient when used on the robotic system has to nd out in the testing and evaluation phase. Namely, the center of motion is now at the other end of the gripper making the overal movement, which now includes scaling as well, considerably dierent than with the manual instruments.
+ Great similarity with current manual instruments + Thin shape can enhance greater precision - Less room for buttons and technical functions
- Results and movement in the robotic system may dier from manual procedure
3.3 Testing and evaluation 3 CONCEPT GENERATION
3.3 Testing and evaluation
Every generated concepts has pro's and con's that were written down in the previous paragraph. However, this combined with a ranked comparison with the requirements does not provide su cient arguments for choosing a winner. Besides, every concept diers reasonably and is strong in their own kind. In order to gain substantiated material that points out the right concept and direction to go to, a practical user test has been set up which can be executed with several stakeholders (e.g. vitreoretinal surgeon/clinician).
The following subparagraphs will give a brief overview on the preperation, execution and the results &
conclusions of this test.
The models of the four concepts used for these tests, can be seen in gure 28. These concepts are equipped with only one button: the coupling button. Reason for this is that the coupling button is the most important input of the gripper. In this way, a primal best location for a button is searched. The participants can then be asked for optional locations for additional buttons.
Figure 28: Physical models including the coupling button used for the user tests.
3.3.1 Theoretical Preparation
Before a plan can be made for the user test, a clear goal has to be de ned. What is the true purpose of the test? And what results have to be substracted from the test? These are important questions that
rst have to be cleari ed.
The main goal of the test is to point out the right design and direction the development process should be taken to. Additionaly, a proper and convenient placement of the inputs/buttons needs to be found.
