Robotic-Assisted Surgical Tools
Introduction
In recent years there have been significant interest and advances in the development of robotic-assisted surgical systems for several different applications. My most recent research work involved the detailed design (including DFM), prototyping, testing, and validation of a unique tool for Trans-Oral Robotic Surgery (TORS), conducted at the Institute for Craniofacial and Cleft Innovation at the Hospital for Sick Children in Toronto, Canada. Of particular interest for my research here was TORS for cleft palate repair [1, 2] in infants.
Cleft palate is a common birth defect wherein the hard and/or soft palate (i.e., the roof of the mouth) does not completely fuse together, resulting in an opening or split (see Figure 1). Complications from this type of defect include difficulties with feeding as well as speech, and even hearing problems. However, this type of defect can be repaired through surgical intervention.
Figure 1: Illustration of a cleft palate defect in an infant (image obtained from [3]).
A cleft palate repair operation is typically done when the patient is between 11 and 14 months of age. Nevertheless, this surgical repair procedure can be difficult due to the following challenges [4]:
1) Precise dissection and interaction with delicate tissues is required.
2) The infant oral cavity presents a very tightly confined workspace within which to perform the surgical tasks.
3) Due to this small workspace, conventional surgical instruments have limited manipulability and limit the surgeon’s visibility of the surgical site.
Compared to conventional surgery using handheld tools, robotic-assisted surgery provides the benefits of improved visualization of the surgical site [5], and the ability to further miniaturize the tools, leading to better access to the workspace [6]. Moreover, the use of surgical tools through a tele-operated, robotic platform equipped with appropriate control software allows for refined control of the tools and, therefore, improved dexterity [7]. This facilitates manipulation of the delicate tissues involved in TORS for infants, and minimizes the potential for trauma to the patient from unintentional collisions between the tool and the oral cavity or from jerky movements when handling delicate tissue. Thus, infant cleft palate repair is an application area that stands to benefit significantly from the implementation of robotic-assisted surgery with specialized tools.
Our TORS tool for infant cleft palate repair was designed and built to function on the da Vinci system platform from Intuitive Surgical, Inc. (Sunnyvale, CA, USA). The da Vinci system is a pioneering robotic-assisted surgical platform that has found increasing adoption within hospitals and research labs and institutes [8]. It is a 3-component system that includes the following [9]:
1) The surgeon’s console, which is the tele-operation user-interface that the surgeon interacts with.
2) The control cart, which houses the computing hardware and software for both, the controller that translates the surgeon’s joystick movements to the actual movements of the tool tips that interact with the patient, as well as the vision system that takes the input from the endoscope and provides it in 3D HD vision to the surgeon.
3) The patient cart, which contains the robot-arms that are equipped with actuators that interface with the different removable/replaceable surgical tools.
The patient cart thus employs a modular design that allows for different types of tools to be attached to any of the robot-arms available on the cart via a standardized tool-actuator interface. As an example of a typical da Vinci system tool, Figure 2 shows a “Classic-generation” model scissor tool. All tools of this model use the same overall design:
-
a body, which houses the capstans onto which the tool-tip-actuating cables are wound, as well as the interface that mates with the actuators on the patient-cart robot arms;
-
a long shaft, along which the actuating cable sub-assembly runs; and,
-
an articulated wrist mechanism with a particular surgical tool tip (e.g., scissors, elevator, needle driver, etc.).
The 4 Degree-Of-Freedom (DOF) articulated wrist on the da Vinci tools provides improved dexterity for working within confined spaces compared to conventional, hand-held endoscopic surgical tools. However, these tools employ an 8-mm-diameter shaft, which are typically used in adults, and most often within the relatively larger abdominal cavity. They become cumbersome to use in infants, and easily clutter the particularly small workspace when operating within the infant oral cavity, especially when using multiple tools simultaneously.
The purpose of this research effort, therefore, was to develop a miniature articulated wrist design for robotic-assisted surgical tools that are optimized to address the challenges for TORS in general, and for robotic-assisted infant cleft palate repair in particular. The design needed to be refined such that its components could be machined by conventional machining and fabrication techniques to produce a usable tool for physical experimentation and validation by conducting surgery on simulated and live tissue.
Since our focus was on mechanism design, we adapted our concept to the da Vinci Classic-generation tool module. By developing our unique wrist-mechanism concept on top of the da Vinci tool module platform, it allowed us to employ the da Vinci Research Kit (dVRK) available in our lab for rapid validation and testing.
The dVRK is a barebones, research version of an older generation, retired da Vinci system, complete with a surgeon’s console, a control cart, and a patient cart with 3 robot-arms. The controller hardware and software, along with a functioning tele-operation interface on the surgeon’s console, and robot-arms on the patient cart, together serve as a readily-available, functional, robotic-assisted surgical platform for testing. This system can easily be configured to control any custom tool that we develop through modifications to adjustable parameters in the controller user-interface program, so long as the physical interface on the tool body remains compatible with the actuator interface on the robot-arms of the dVRK. Experimentation is therefore streamlined and cost-effective, allowing us to accelerate the iterative design cycles.
Design Overview
Since we prototyped our design on top of the da Vinci tool module platform, we scavenged the tool bodies from old, used Classic-generation tools obtained by permission from our contacts at Intuitive Surgical, Inc., and modified them to incorporate our articulated wrist. Figure 3 shows an example of our custom tool design with a needle driver tip.
Key aspects of our tool design to note include:
1) A compact, 4-DOF, cable-driven, articulated wrist design that fits within a 3mm-diameter envelope.
2) A two-stage shaft comprised of a 5mm-diameter segment extending out from the tool body and a 3mm-diameter segment leading to the wrist components.
3) A custom-designed, pitch-to-yaw decoupling mechanism incorporated into the tool body.
Figure 3: Example of custom, 3mm-diameter tool with needle driver tip.
The following sub-sections briefly provide a closer look at some of these key aspects of our tool design.
The Articulated Wrist
The articulated wrist is a 4-DOF, cable-driven mechanism composed of 4 pin-jointed links. Figure 4 shows an exploded-view CAD model of the links that form the wrist sub-assembly.
Figure 4: Exploded view of 4 links comprising the articulated wrist design (needle driver tool tip shown here).
The key design concept that enables the miniaturization of the articulated wrist to within a 3mm-diameter envelope is the use of static grooves and guide-channels that are machined into the links themselves (see Figure 5). This eliminates the need for any additional components for cable guidance within the wrist sub-assembly such as pulleys and idlers, thereby decreasing the overall number of components required and reducing the size and length of the links. The shorter links allow for compact articulation of the wrist, and their smaller overall size results in less obstruction when operating within confined workspaces.
Figure 5: Close-up view of the 3mm-diameter wrist links showing the machined grooves and guide-channels for the actuating cables.
The disadvantage of using grooves and cable guide-channels is increased friction and, consequently, greater cable tension required when actuating the links. Nevertheless, this was expected, and an earlier study that we conducted on the characterization of this increase in cable tension due to friction as a function of wrist pitch showed that the net tension remains safely below the cable’s failure threshold. Moreover, physical tests with the fully-assembled tool prototype on an infant cleft palate phantom model with simulated tissues showed that the cable sub-assembly had sufficient strength to carry out the required surgical tasks despite this increase in cable tension due to overcoming friction.
The Decoupling Mechanism
The articulated wrist is separated from an actuation body by a 500mm-long steel shaft sub-assembly. In particular, Link 1 of the wrist sub-assembly attaches onto a 3mm diameter steel tube, which in turn connects to a 5mm diameter tube (for greater rigidity of the overall tool shaft) through a 3D-printed shaft adaptor. The actuating cables that pass over the grooves and guide-channels on the links run the length of this shaft and into the tool body.
The tool body contains the sub-assemblies that:
-
manage the wrapping and un-wrapping of the cables to engage the different DOFs;
-
coordinate the yaw and grip motions relative to pitch; and,
-
interface with the actuators on the dVRK robot-arms on the patient cart.
A significant addition to the standard Classic-generation tool body has been a custom-designed mechanism to decouple the pitch motions from the yaw and grip DOFs. This addresses the coupled nature of the pitch and yaw/grip DOFs inherent in this cable-driven articulated wrist design, wherein movements about the pitch axis result in changing path lengths for the cables that actuate links 3 and 4 to achieve the yaw and grip positions. As such, slack in these cables must be added or removed accordingly to maintain cable tension and, thus, the desired yaw and grip positions, as pitch changes.
Figure 6 gives an overhead, close-up view of the decoupling mechanism incorporated into the tool body, both fully-assembled (left image) and with some of the top covering components removed (right image). The mechanism comprises custom-profile cams that interact with idlers attached onto freely-moving platforms on linear guide rails. In addition, the cables that control the yaw/grip motion wrap around different pulley-sets, some of which are attached onto the tool-body base and two of which are attached onto the linear guide-rail platforms (see cable paths in blue in Figure 6). The cam rotation is coupled to pitch rotation (see cable paths in red in Figure 6), with the cam profiles designed in such a way as to push the platforms on the linear guide rails by the correct amounts to maintain tension in the cables that control the yaw/grip motion during wrist pitch actuation. The tension in the yaw/grip cables, in turn, ensure that the idlers on the sliding platforms always maintain contact with the surface of the cams.
Figure 6: A closer view of the tool body showing the components that comprise the pitch-to-yaw/grip decoupling mechanism and associated cable paths.
The cams were designed by determining the analytical relationship between the wrist pitch angle and the length of the cable that exits from Link 2 and runs to the base of Link 1. The cam profile was determined by computing the required radius at different angular positions of the cam to appropriately translate the sliding platforms so as to compensate for the corresponding change in path length (see [10] for the detailed mathematical computations).
The Working Prototype
Figure 7 shows an overhead view of the overall, final 3mm tool prototype. Figure 8 shows a close-up of examples of the assembled, articulated wrist sub-assembly with 3 different tool tips, namely, a needle driver, scissors, and a grasper.
Figure 7: Overall 3mm tool prototype for robotic-assisted infant cleft palate repair.
Figure 8: Closer view of the 3mm tool articulated wrist sub-assembly with 3 different tool tips (from left to right: needle driver, scissors, and grasper).
Intensive testing and validation were also conducted using these tools with the dVRK and a physical, high fidelity, cleft palate model. Figure 9 shows different views of the test setup. The central robot arm holds the 3DHD endoscopic camera that relays visual feedback to the surgeon at the tele-operation console, and the two robot arms on either side hold the two different tools that the surgeon chooses to use. The vertical alignment of the endoscopic camera over the oral cavity opening of the cleft palate model can be changed according to the surgeon’s preference for better visualization during the different steps of the surgery. Figure 10 demonstrates the improved visibility afforded by the 3mm tools inside the oral cavity of the model (as seen by the surgeon from the endoscopic camera) compared to the standard 8mm tools that are commercially available on the da Vinci systems.
Figure 9: Illustration of the experimental setup for tool prototype validation: a) overview of the setup showing 3 robot arms of the da Vinci Research Kit (dVRK), with central arm holding the Endoscopic Camera Manipulator (ECM), and 2 arms on either side holding tools with 2 different tool tips; b) experimental setup with ECM at a 10° angle relative to the vertical for better visualization by the surgeon during a particular surgical step; and, c) experimental setup with ECM aligned vertically for midline surgical steps.
Figure 10: Surgeon’s view from the endoscope camera showing visibility within the infant oral cavity of the cleft palate simulator when using a) the da Vinci 8mm diameter, and b) the prototyped 3mm diameter tools.
Each validation experiment involved conducting several iterations of suturing tasks to re-approximate the separated nasal and oral mucosa of the simulated cleft defect in the model. These tests were used to assess the ease with which the surgeon was able to reach the different key regions of the workspace, as well as the overall feasibility of physically performing a complex TORS procedure using the 3mm tool prototypes. Performance measures included the number of unintended collisions with the oral cavity tissues per suture (which gauges the potential for unnecessary trauma in practice), percent occlusion of the surgeon's field of view by the tools, and the percent of kinematically feasible poses that are collision-free for an array of target points defined within the workspace (for the computer-simulation-based reachability analysis). Experiments were also repeated for different tool trocar positions in order to gauge the impact of this parameter on these performance measures as well.
Given below is a short video showing part of a suturing task that was conducted during the physical experiments using the 3mm tool prototypes. On the left is a grasper tool and on the right is a needle driver. The video shows the visual feedback from the 3DHD endoscopic camera of the roof of the mouth inside the highly restricted workspace of the infant oral cavity within the cleft palate model (top is back of the mouth; bottom is the front). Lateral relaxing incisions can be seen placed through the oral mucosa on either side of the midline, and the surgeon must re-approximate the tissue along the midline through several sutures.
Video 1: Surgeon performing part of a suturing task on an infant cleft palate physical simulator model using the 3mm tool prototype with the dVRK.
More details on the physical and simulation experiments that were conducted for tool validation can be found in the corresponding journal publication that resulted from this work in the IEEE Transactions on Biomedical Engineering, entitled “Robotic Assisted Cleft Palate Repair Using Novel 3 mm Tools: A Reachability and Collision Analysis” [11].
[1] D. N. Nadjmi, "Transoral robotic cleft palate surgery," The Cleft Palate Craniofacial Journal, vol. 53, no. 3, pp. 326–331, May 2016.
[2] K. Khan, T. Dobbs, M. C. Swan, G. S. Weinstein, and T. E. Goodacre, "Trans-oral robotic cleft surgery (TORCS) for palate and posterior pharyngeal wall reconstruction: A feasibility study," Journal of Plastic, Reconstructive & Aesthetic Surgery, vol. 69, no. 1, pp. 97–100, Jan. 2016.
[3] "Cleft lip and cleft palate inner" [PNG Image]. (n.d.). Vikram ENT Hospital. https://www.vikramenthospital.com/blog/cleft-lip-and-cleft-palate.php.
[4] S. Vadodaria, N. Watkin, F. Thiessen, and A. Ponniah, "The first cleft palate simulator," Plastic and Reconstructive Surgery, vol. 120, no. 1, pp. 259–261, Jul. 2007.
[5] J. C. Jacobson, M. G. Scrushy, L. A. Gillory, and S. R. Pandya, "Utilization of robotics in pediatric surgical oncology," Seminars in Pediatric Surgery, vol. 32, no. 1, pp. 151–263, Feb. 2023.
[6] G. Pelizzo, G. Nakib, P. Romano, L. Avolio, S. Mencherini, E. Zambaiti, A. Raffaele, T. Stoll, N. Mineo, and V. Calcaterra, "Five millimetre-instruments in paediatric robotic surgery: Advantages and shortcomings," Minimally Invasive Therapy & Allied Technologies, vol. 24, no. 3, pp. 148–153, Jun. 2015.
[7] J. C. Jacobson and S. R. Pandya, "Pediatric robotic surgery: An overview," Seminars in Pediatric Surgery, vol. 32, no. 1, pp. 151–255, Feb. 2023.
[8] E. I. George, T. C. Brand, A. LaPorta, J. Marescaux, and R. M. Satava, "Origins of robotic surgery: From skepticism to standard of care," Journal of the Society of Laparoendoscopic Surgeons, vol. 22, no. 4, Oct.-Dec. 2018.
[9] "da Vinci Xi system brochure," Intuitive Surgical, Inc., Jun. 2019, www.intuitive.com/en-us/-/media/ISI/Intuitive/Pdf/da-vinci-xi-system-brochure.pdf. Brochure.
[10] G. C. Y. Wu, D. J. Podolsky, T. Looi, L. A. Kahrs, J. M. Drake, and C. R. Forrest, "A 3 mm wristed instrument for the da vinci robot: Setup, characterization, and phantom tests for cleft palate repair," in IEEE Transactions on Medical Robotics and Bionics, vol. 2, no. 2, pp. 130–139, May 2020.
[11] G. Maguire, E. Tang, T. Looi, and D. Podolsky, "Robotic assisted cleft palate repair using novel 3 mm tools: A reachability and collision analysis," in IEEE Transactions on Biomedical Engineering [accepted; in press].