gms | German Medical Science

The development of an interactive robot system is motivated by middle ear surgery as an example of microsurgery. In this scope of surgery the surgeon is often confronted with very small and sensitive structures like the ossicles (hammer, anvil and stirrup), nerves (nervus facialis and chorda tympani) and vessels. The milling process is an essential part of the intervention because it is necessary for widening the operation field as well as for cleaning otosclerotic regions. Thereby the main problems are:

1.

There is a conflict between the tremble free positioning or manipulation of the fine structures of the middle ear or a prosthesis and the foregoing exhausting milling process to get access to the middle ear [4].

2.

The milling instrument can slip of and cause an injury of the ossicles, the vessels or the nerves [11] like a noise trauma, luxation of the ossicles or the like. This can produce temporary or permanent hearing loss.

3.

Minor failure during the manipulation can even cause a severe injury of the ossicles, the vessels or the nerves. Hurting the nervus facialis or the chorda tympani can lead to a loss of the sense of taste and salivation.

These problems could lead into complications as described above or cost-intensive following surgeries.

State of the art for supporting surgical interventions is computer-aided surgery (CAS). Systems of CAS can be further divided into navigation systems and robot systems. Most navigation systems underlie the same general principle. With the aid of a planning component the surgeon can define target positions within the coordinate system of the preoperatively taken CT or MRI data set of the patient. The position of a surgical instrument is measured relative to the patient during the intervention. The navigation system guides the surgeon to the defined target position and provides visibility during poor visibility in reality. A commercial navigation system for the usage in ear, nose and throat surgery (ENT) is for example Vector Vision ENT by BrainLab. The disadvantages of navigation systems are that tremor can not be reduced and a slipping off can not be prevented.

Robotic systems can be differentiated by their control principles [8] in automatic systems, telemanipulative systems and interactive systems.

1) Automatic systems execute specific pre-planned surgical intervention steps like the driveaway of a defined position or the guidance of an instrument along a trajectory automatically. The systems can even control the velocity or force. The specific surgical intervention step is fulfilled automatically but always with the possibility of interrupting it. Examples in this field are the Robodoc (ISS, Sacramento, USA) and the Caspar (Orto Maquet, Germany) system. A system concerning milling operations is the medical robotic system RONAF [10]. Disadvantages of all these automatic systems are the limited capability of intervention, especially in unpredictable intraoperative incidents. Another disadvantage is that they are basically built up out of industrial robots which are adapted for medical applications. These systems are quite large which results in a restricted work area for the surgeon.

2) Telemanipulative systems have no planning component. They do not provide any positioning control. The intervention is done by the robot system, which is direct controlled by the surgeon via special human-machine-interfaces. Thereby the surgeon is supported in microsurgical interventions by scaling down forces and increasing the haptic feedback. For this a direct or indirect sight on the field of intervention is necessary. Telemanipulative systems are for example the daVinci (Intuitive Surgical, Sunnyvale, USA) and ZEUS (formerly Computer Motion Inc. now Intuitive Surgical, Sunnyvale, USA) system. A system developed for middle ear interventions is the Steady Hand Surgical Assistant System [1].

3) Interactive systems combine the advantages of automatic and telemanipulative systems. They provide a preoperative planning component and position control as well as a direct interaction with the surgeon. This user interaction is an important point. If the direct control input of the user is missed, some systems interpret the absence of these signals as a stop signal. With this mechanism an uncontrolled movement can be prevented. Interactive systems are for example the Acrobot [5], neuroArm [9] and the first clinical applied interactive robotic system SurgiScope 2000 [3]. For solving the medical requirements of the focused microsurgery inventions such interactive systems are most capable.

For preventing slipping off during the milling process the interactive robotic system MicroAssistant is designed. The MicroAssistant (Figure 1 [Fig. 1]) is in preclinical evaluation but addresses clinical applications in the field of microsurgery. The system was first evaluated as an automated system and is described in [7]. The MicroAssistant contains a planning component, an interactive robot system and an optical pose measurement system. Higher stiffness of the whole system is obtained through a parallel kinematic structure, which is a new approach and detailed described in [2] and [6]. The system only allows a movement along a plane (x-y axes) which supports the lateral milling. It has to be positioned manually above the operation field. For guidance and positioning of an instrument the optical pose measurement system MicronTracker2 (Clarontech, Toronto, Canada) is embedded. The control of the system is fulfilled through a new kind of human-machine-interface which is described in detail in the next section. To evaluate the control principle and human-machine-interface the MicroAssistant is actually handled as a telemanipulative system.

Any sort of interaction with a robot system needs a human-machine-interface (HMI). The new approach for the design of the human-machine-interface as a joystick with three degrees of freedom is derived from the aim to disturb the surgeon’s actual habits as less as possible. Therefore a kind of joystick is developed which can be used with the heel of the hand (Figure 2a [Fig. 2]). By doing so it is still possible to hold the clamped instrument with the fingertips while giving control signals via the human-machine-interface. The human-machine-interface contains a hand and a base panel. The base panel is rigidly attached at the MicroAssistant and connected with springs to the hand panel (Figure 2b [Fig. 2]). The user can apply forces to the hand panel through his heel of hand. These forces result in a relative dislocation of the hand panel to the base panel.

One marker of the optical pose measurement system MicronTracker2 (Claron Technology, Toronto, Kanada) is attached on both panels of the human-machine-interface, because the whole control principle based on the pose measurement system only. The positions of the hand and base panels are measured during the control process (). Two reference coordinate systems (r ₁ , r ₂ ) are defined at the juncture of the hand and base panels (Figure 3 [Fig. 3]), because this point is the origin of each deflection of the human-machine-interface. One reference coordinate system is defined for the hand (r ₁ ) and one for the base panel (r ₂ ). If the human-machine-interface is in rest position, the origins of the two coordinate systems lie on top of each other. If the human-machine-interface is in use the origins are displaced.

To calculate the relative displacement of these reference coordinate systems a calibration is needed to get the transformation matrices () from the markers (hand, base) to the depending reference coordinate system (r ₁ , r ₂ ). The displacement can be calculated as coordinate transformation with the aid of the pose measurement system as follows:

(1)

The position part () of the matrix is the desired relative difference vector between the hand and base panel. This displacement can be interpreted and used as two dimensional control vector for the robotic system.

The third degree of freedom is the rotation of the hand panel relative to the base panel. The rotation a is the angle between the two x-axes of the reference coordinate systems r ₁ and r ₂ . This rotation can be calculated with the arccosines (2). The rotation matrix already contains the rotation in the x-y-plane because the human-machine-interface is restricted to this plane.

(2)

The third degree of freedom is not used yet in the MicroAssistant controlling scheme. If the movement mode of the MicroAssistant is disabled, the translational displacement of the human-machine-interface can be used to move the camera position of the visual scene or the sphere surface. The rotation of the human-machine-interface can be used to zoom the visual scene. Because the human-machine-interface is only based on the pose measurement system there is no need for extra sensor technology and electronics assembly. Therefore the flat design of the human machine-interface is possible and there is no electronics assembly near the patient.

Milling with the aid of the MicroAssistant under control of the new human-machine-interface can solve the conflict of slipping of during the milling process for widening the operation field. But the new concept constrains the surgeon to rearrange the milling process, because the instrument is clamped at the robot system and its orientation can only be changed by repositioning of the MicroAssistant. Because of that the milling instrument could obstruct the direct sight onto the scene. The surgeon is constricted to a slide by slide milling order too, because a milling in depth has to be adjusted manually. To test whether the MicroAssistant with control via the new human-machine-interface is although usable for supporting the milling process the accuracy of the freehand milling is compared with the accuracy of the robot assisted milling. Also effects of the system’s dynamic which can cause a latency time between the user’s input and the movement of the robot system are analyzed as well as the noise of the pose measurement system that can lead to undesired movements in any direction. The experiment is described in detail in the next section.

For testing the control usability of the joystick the accuracy of the freehand milling is compared with the robot assisted milling. A predefined reference trajectory is applied on a plaster model. This trajectory should be milled inside the plaster model by using the human-machine-interface (Figure 4a and 4b [Fig. 4]) and doing it manually (Figure 4c [Fig. 4]). For the decision in which direction the subject has to mill only visual feedback of the applied trajectory is provided and allowed. To compare the results with the guidelines the plaster model with the predefined trajectory is scanned before the milling process and after the milling process. The scanned images are compared and the differences are calculated at 17 points along the trajectory. To get a statement whether the joystick is suitable the experiment has been done by 3 different subjects.

The reference trajectory with the points of comparison is depicted in detail in Figure 5a [Fig. 5]. The scanned applied trajectory before the milling process is shown in Figure 5b [Fig. 5]. Altogether a total of 68 measurement points were taken for each experiment. For freehand milling the maximum error is 0.71 mm and the mean error is 0.19 mm with a standard deviation of 0.16 mm (Figure 5c [Fig. 5]). The robot assisted milling process leads to a maximum error of 0.67 mm and a mean error of 0.12 mm with a standard deviation of 0.17 mm (Figure 5d [Fig. 5]).

The results show that the robot assisted milling is as accurate as the freehand milling because the maximum error of 0.71 mm for the freehand milling and the maximum error of 0.67 mm for the robot assisted milling are nearly the same. The same is applying for the mean errors and the standard deviations. This fact shows that system’s dynamic and the noise of the pose measurement system do not lead to latency times or affect the accuracy of the milling.

It is conspicuous that the mean error is out of all proportion to the standard deviation in both cases: the freehand milling and the robot assisted milling. This could be due to the fact that the milling instrument or the plaster dust is obstructing the direct sight onto the reference trajectory. The same situation can appear in real surgical interventions caused by bodily fluids or the like. An integration of the planning component of the system could lead to better results and will be tested in subsequent work.

The MicroAssistant is an interactive robot system for computer aided surgery that could minimize complications in middle ear surgery as an example of microsurgery. It could prevent slipping of during the milling process for getting access to the field of surgery. The system is controlled by a new kind of human-machine-interface without any additional sensor technology and electronics components. Hence there is no extra electronics near the patient. Although there are constrictions to the surgeon milling with the aid of the robot system is as accurate as freehand milling.

Conflicts of interest: none declared.