Development of a Robot-Assisted Surgery System for Cranio-Maxillofacial Surgery

Medical robots have been developed rapidly in recent years. Clinical application of da Vinci system showed its advantages. Currently, there is no specialized robot system for cranio-maxillofcial surgery. We developed a cranio-maxillofacial surgical robot system focusing on the reconstruction of mandibular defects.


General Information of Surgical Robots
Surgical robots have been tried in many fi elds of surgery like in orthopedics, microsurgery, neurosurgery, ENT surgery, catheterization procedure, etc. The fi rst robot assisting orthopedic operations was made by Integrated Surgical System Company, called ROBODOC system, in 1992 [ 1 , 2 ]. It was developed from industry robots and was able to perform operations like knee joint replacement. Since then the other robots assisting surgery were developed, but very few of them could be used in clinic. In August of 2015, a newly developed robot by a research team in Beijing was used to assist spine surgery in a real patient and got satisfactory result. Currently the commercially available surgical robot is da Vinci surgical robot which helped in fi nishing prostatectomy for the fi rst time in the University of Frankfurt in 2000 and has become the main and much welcome surgical robot in the world. So far, as we know, there is no specially designed surgical robot for cranio-maxillofacial (CMF) surgery [ 3 -6 ]. We began to develop our system in 2010 and have made two robots: one for mandibular reconstruction and the other for needle insertion to diagnose and treat lesions in the skull base. In this chapter we introduce the fi rst robot system.
Our robot-assisted surgery (RAS) system for CMF surgery consisted of the following three parts: computer-aided design (CAD) system for CMF surgery, surgical robot for CMF surgery, and medical experiment for the CAD and robot. The outlines of RAS system for CMF surgery is showed in Fig. 5

Computer-Aided Design (CAD) System for CMF Surgery
The CAD was designed for four typical and more or less diffi cult operations: orthognathic surgery, reconstruction of the orbital defects, resection of skull base tumors, and reconstruction of mandibular defects. It has the following fi ve main functions: volume rendering of medical images, segmentation on medical images, CT/MRI image fusion, craniofacial database service, and virtual surgery designing.
The software is easy to use with good interactive regulation of transfer function. It can produce high-quality volume rendering of medical images and get quick and precise segmentation of soft tissue tumors based on graph-cut algorithm. Abstracted tumor from CT data can be visualized three dimensionally and measured for its volume with 99.5 % agreement of the true volume of the tumor. GPU marching cubes algorithm was used to build highly precise three-dimensional reconstruction of bony structures of the skeleton from CT images. The CT/MRI image fusion is another character of the CAD; it well fuses the bony structures abstracted from CT with the vessels abstracted from MRI and creates a new image with more information for more precise medical image analysis and treatment planning. It can also fuse other images obtained by the other image equipments, like fusion of CT data with 3D laser scan image data for the dentition.
We also built a CMF database with more than 20,000 CT data of CMF to assist preoperative design. Currently, one of its functions we already used in clinic is to search for "similar skull." We selected 71 landmarks on the CMF bones. Based on the landmarks, it can retrieve "similar skull" in a few seconds.
With the above mentioned basic functions, the CAD performs its most important task: carrying virtual surgery design for different operations, mainly for orthognathic surgical management, mandibular reconstruction, orbital reconstruction, and removal of the skull base tumors. The designing system is easy to be learned and used for virtual surgical planning step by step. All procedures can be stored and repeated, which helps young surgeons to learn how to design and do these operations. After the virtual planning is done, 3D stereolithographic model (STL model) can be manufactured from the virtual planning model by using rapid prototyping technique for model surgery. Or the virtual planning data is inputted into the navigation system to guide surgeons or robots to perform operations according to the virtual preoperative plan. Thus the precise real-time registration of the CAD system with navigation system is the key point for realization of accurate navigation.

Surgical Robot for CMF Surgery
The development of the robot had the following main contents: conformation of robotic system, control system, safety control of the robotic system, navigation and trail program, and test of the robot.

Conformation of the Robotic System
The robot had three arms and six degrees of freedom, its hold strength was 30-50 N, and its working square was 200 × 200 × 200 mm. Each arm consisted of two segments and three joints. Figure 5.2 shows the mechanical structure of the robot and the appearance of the robot after assembling.

Safety Control of the Robotic System
Safety is the fi rst consideration for clinical application of medical robots. It was reported that more than 60 % of adverse events in clinical use were caused by robots' malfunction. Our robot safety control involved the workfl ow from patients' diagnosis, virtual surgical planning, navigation to robot assisting surgery, robotic system itself like essential electro-circuit for safety control, interactive process, and

Navigation and Trail Program
Navigation was realized by coordinate switch of four coordinates: 3D object space, image coordinate {V}; physical space, patient coordinate {P}; robotic space, robotic coordinate {R}; and optical space, optical coordinate {M}. The algorithms of trail program included cube polynomial and quintic polynomial.
After all these works were completed, the whole system of the robot was packaged. It consists of the robot, optical tracker, and workstation.

Test of the Robot
We assessed positioning accuracy of the robot and found its repeat positioning error was less than 0.10 mm and the systemic absolute positioning error <2.45 mm, which met the requirement of the design indicators for performance check. And the robot was ready for medical experiment.

Evaluation of the CAD Software
Twenty-fi ve young oral and maxillofacial surgeons were asked to use and evaluate the CAD software. The satisfaction degree was 91 %.

Accuracy of the Three-Dimensional Reconstruction of the CAD Software
The accuracy of three-dimensional reconstruction and the accuracy of the navigation system and the robot were all tested by model or animal study. Thirty model skulls were used to test the accuracy of the three-dimensional reconstruction of the CAD software. Titanium screws of 0.1 mm were fi xed on the models. The marked models were CT scanned. Reconstruction of the models was performed on the CT data. The comparative study between the reconstructed models and the real models showed that the error was 0.21 mm.

Accuracy of the Navigation System
For navigation assessment, the fl owchart of the animal experiment includes the following: (1) use living goats as study animal, (2) fi x marks on the cranio-maxillofacial region, (3) have CT scan, (4) have virtual surgical design, (5) do operation under navigation, (6) have postoperative CT scan, and (7) compare pre-and postoperative results. All of the four typical operations were tested for navigation accuracy. Comparison of postoperative results with preoperative design showed that the error of mandibular reconstruction was 2.9 mm, error of orthognathic surgery 2.7 mm, error of orbital reconstruction 1.2 mm, and error of skull base surgery 1.8 mm.
There are many factors that can affect the accuracy of navigation. These factors are CT scan slice thickness, resolution and radiation dosage, shadow of titanium plates, accuracy of three-dimensional reconstruction, titanium plate bending, movements of remaining bone segments, etc. Current animal study results are acceptable for clinical application.

Animal Experiment of Robot-Assisted Operations
The experimental fl ow was similar to the abovementioned navigation study. After the animal CT data were collected, the preoperative design was planned using the CAD system. The robot and animal registration was performed, the surgery of making osteotomy on the mandible was done, and the fi bular bone graft was fabricated under navigation according the preoperative design (Fig. 5.3 ). When the mandibular defect was made, the robots' two side arms held two remaining bone segments, the The robot animal study showed that the robot can move smoothly and do accurate placement of bone graft and aid fi xation. Its three arms can coordinate well to aid bone graft placement for different types of mandibular defects. Its fl exibility, bone holding method, miniaturization, etc. need improvement for higher stability and accuracy.
Open Access This chapter is distributed under the terms of the Creative Commons Attribution 4.0 International License ( http://creativecommons.org/licenses/by/4.0/ ), which permits use, duplication, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the work's Creative Commons license, unless indicated otherwise in the credit line; if such material is not included in the work's Creative Commons license and the respective action is not permitted by statutory regulation, users will need to obtain permission from the license holder to duplicate, adapt or reproduce the material.