Concept and Design of a Hand-held Mobile Robot System for Craniotomy

Gavin J. Kane
Within the framework of the European Project, Computer Aided Training for Surgeons and Engineers (CompuSurge) was a solution for minimal impact robotic assisted surgery developed [58] [57]. Problem Background Craniotomies, surgical procedures involving the opening of the skull, are a common medical option for the correction of a number of medical diseases and problems in Neurosurgery as well as other areas such as Maxillo-Facial Surgery. It can be required as remedy for Intra-cranial bleeding,
more » ... -cranial bleeding, or to allow surgeons access to intracranial tumours. Additional reasons for performing a craniotomy can include allowing surgeons to insert probes into the brain for treatments of diseases such as Parkinsons, or to assist in the repair of the skull and other cranial problems after trauma. One final reason for performing a Craniotomy is in the treatment of Craniosyntosis. Craniosyntosis, the premature fusion of one or more sutures of the skull, is a rare congenital defect. The standard treatment for a diagnosed case of Craniosynostosis is a surgical remodeling of the skull. In the majority of cases this involves the Fronto-orbital advancement, but can require more complex strategies up to a complete reshaping of the calvaria. The planning for this surgery is based on two significant factors, allowing the brain to grow again normally without any hindrances, and for aesthetic reasons. One significant problem within the surgical procedure is the conveyance of this planning to the patient. The current method involves the surgeon remembering the plan, and using his / her experience drawing fresh the plan on the skull. This method v Abstract is clearly not accurate, and a second problem is that prior to the craniotomy occurring, this plan can be significantly smudged and distorted. There have been many attempts at performing robot assisted surgeries for this procedure; however, most of them involve the adaption of large industrial robots that impact significantly on the operating room, the surgeon and the surgical workflows. They have been poorly accepted by the surgeons because in this critical section of the procedure, their years of experience are wasted as they only watch. Aim of the Project To solve the outlined problems, this work aimed to develop a completely new concept for robotic surgery. A new robot was engineered that was small, mobile and hand-held, capable of seamless integration into the Operating Room with the exact same accuracy of the larger industrial modified robots. In this thesis, I describe the work involved in the system architecture definition, mechatronic design, implementation of the electronic drives, navigation algorithms, software GUI for the surgeon and ancillary safety watchdog systems. System Architecture and Workflow Integration The robot was realised as a hand-held mobile robot, with two active driving wheels for navigation along the trajectory. The robot allowed active controlled motion along a trajectory that was pre-operatively planned on a separate workstation. The Surgical Craniotomy Tool is fixed between the two wheels and cuts directly along the trajectory. The position of the robot was tracked passively by an optical tracking system. The surgeon holds and handles the robot like a normal tool, with the speed intuitively controlled through the tilting of the robot. vi Abstract Mechatronic Design The first robot prototype was developed and built with two actively driven wheels, and passive control over the cutting depth. This was later further developed to include two more active axes for cutting depth and cutting angle. The robot has a complete weight of only 800gms. The motor section may be removed and after sterilisation reconnected with a self-aligning mechanism. This concept protects the sensitive electronic components and encoders from the heat and moisture during sterilisation. The cutting angle and depth is controlled through a parallel kinematic linkage of the wheel axles. The ability to cut on an angle prevents a disjointed section of bone from being pushed into the cranial cavity and pushing on the brain. To ac hieve the necessary traction on the wheels, an opposing force is generated on the underside of the skull by a hook at the end of the craniotomy tool. The wheels are fitted with spikes which can penetrate the top layer of bone, and maintain grip despite the use of water spray. The wheel design includes a security flange preventing excess penetration and allowing their use on the softer thinner skulls of infants. Navigation Algorithms and Man-Machine Interfaces The movement of a mobile robot on a 3 dimensional surface has found very little attention in previous approaches of robot-assisted surgery. In this application, the robot and trajectory segments are tracked in all 6 degrees of freedom. For the navigation of the hand-held robot only three degrees of freedom are required. The use of a Frenet Frame allows this necessary reduction for the control loop without any negative side effects. The tilt of the robot, which is used to control the speed, shifts the cutting axis of the Craniotomy relative to the contact point of the wheels. The same is true for the lateral lean of the robot due to the skull curvature. There variations need to be compensated. This compensation is achieved with adjustment of the Frenet Frame definitions. This Frenet Frame is then used as input vii Abstract to a velocity control loop for the active steering of the robot in minimising the angular error. The robot's velocity is then controlled by the tilt of the robot. The robot tilts around it's wheel axis when the surgeon pushes or pulls intuitively. To determine the inclination, the position upon activation of the drill through its footpedal is used as a reference for each cut. Safety System Integration The risk analysis had identified two significant events that could lead to a substantial risk that could not be mitigated within the initial system structure and workflow. The first problem related to a poor registration, and the second was a potential software error. It is very difficult to improve on the registration accuracy of the point to point method used, therefore additional data was collected to allow assessment of the likely validity of the registration. After the point to point registration, the surgeon uses the pointer to trace out the edge of the skin flap on the skull. This data is used for a second registration through the ICP algorithm with the CT data. This registration is not used for navigation, but the correlation between these two registration methods provides an indication to the surgeon on the likelihood of validity. Additionally this data is used as an absolute safety border for the robots navigation. During the operation, further data is then gathered from the robot, updating the ICP algorithm, and maintaining an updated validity check of the registration. The second problem is more serious, that is a software error leads to an interruption in communication, the PID-controlled motors will continue to drive past any safety border at the velocity last set before the communications interruption. To solve this problem, a new Safety Watchdog was implemented on an FPGA to statistically monitor the communications in the system. If any abnormality is detected the motors can be shut off. viii Abstract Conclusion This work described a new robot assisted surgical system that could be completely integrated into the surgical workflow that in contrast to previous approaches is highly intuitive to use. The system allows the cutting of accurate craniotomies, based on pre-operative planning and was achieved without the normal drawbacks of robot assisted surgery, specifically without any large influence on the surgical workflow or impact on the layout in the operating room. The system does not attempt to replace the surgeon, but alternately supports them in the critical phase of the operation. ix Kurzfassung Im Rahmen des Computer Aided Training for Surgeons and Engineers Projekts (CompuSurge) wurde eine Lösung zur minimal traumatischen robotergestützten Chirurgie entwickelt [58] [57]. Problemstellung Kraniotomien, chirurgische Eingriffe zur Öffnung des Schädels, werden häufig in der Neurochirurgie sowie in der Mund-Kiefer-und Gesichtschirurgie durchgeführt. Mit dieser Methode können intrakranielle Blutungen behandelt werden und es wird ein Zugang zu intrakraniellen Tumoren ermöglicht. Desweiteren lassen sich mit Hilfe der Kraniotomie Sonden in das Gehirn einführen, um Krankheiten wie Parkinson zu behandeln oder die Reparatur des Schädels und anderer kranialer Probleme nach einem Trauma zu unterstützen. Außerdem kann mit Hilfe der Kraniotomie Kraniosynostose, die vorzeitige Verknöcherung einer oder mehrerer Schädelnähte, behandelt werden. Die Standardbehandlung für eine diagnostizierte Kraniosynostose ist eine chirurgische Umgestaltung des Schädels, in der Mehrzahl der Fälle durch das "Frontat-orbital Advancement". Außerdem können auch komplexere Maßnahmen bis zu einer kompletten Neugestaltung der Schädeldecke durchgeführt werden. Die Operation hat hauptsächlich zum Ziel, dass das Gehirn wieder normal und ungehindert wachsen kann. Ein zweiter nicht unwesentlicher Faktor ist die Ästhetik. Die Operation wird vom Chirurg mit Hilfe einer sogenannten Kephalometrie-Analyse basierend auf Röntgenbildern oder CT-Aufnahmen xi
doi:10.5445/ksp/1000034461 fatcat:bupqhtnppzhlnldy4nmcsklzvm