Discussion - Robotized Transcranial Magnetic Stimulation - page 150

Furthermore, the presented method for online calibration between robot and
tracking system for navigated robotized systems in medical applications, in par-
ticular robotized TMS, has therefore three major benefits:
• First, it increases the usability of the robotized system for medical users and
therefore the acceptance in the clinical workflow: For standard hand-eye cali-
bration, a marker must be mounted to the robot end effector, the robot with the
marker has to be moved in line of sight of the tracking system, the hand-eye
calibration has to be performed with a set of data points, and the marker has to
be removed, and the tool (e.g. the TMS coil) has to be attached again
(cf. Sect. 4.1 ). This is obstructive in a clinical workflow. The presented robust
real-time robot/camera calibration method, on the contrary, does not require
such a time-consuming calibration step. In fact, computation of the calibration
method is directly performed in less than 200 ms.
• Second, it solves a severe safety issue for navigated robotized systems in
medicine. Without recognition by the user, a movement of tracking system or
robot during the application could be a serious safety issue in a standard setup.
Due to mispositioning with a wrong calibration, the robot could hit the patient or
lab equipment. With the online calibration approach, we continuously check the
calibration during treatment. Thus, we can determine if tracking system or robot
have been moved during the application. In general, it would also be possible to
automatically update the calibration during the application.
• Third, it can easily be adapted for other medical robotized systems. They are
increasingly merit in surgery, treatment and diagnosis to allow for more pre-
cision, accuracy, and reliability. The system designs differ, but many medical
robotic systems require tracking, e.g., of patient movements or target position.
For instance, a surgical robotized microscope uses a tracking device to display
preoperative data in the actual microscope's field of view [ 7 ]. The integration of
a tracking camera to a mobile robotized C-arm supports navigation and aug-
mented reality features [ 4 ]. The presented online calibration can easily be
adapted to these other partially mobile system designs where the Denavit-
Hartenberg (DH)-parameters are known and at least one link is always visible
for the tracking device.
9.2 Hand-Assisted Positioning
We have presented a method that allows the user to move the robot in a hand-
guided fashion. By grasping the TMS coil by hand and acting on the coil, the
occurrent forces and torques are measured with an Force-Torque (FT) sensor.
These forces and torques are consequently transferred into robot movements.
In this way, the user can position the TMS coil with the robot in an intuitive
fashion. Skills on robot control are thus not required for an effective application.