Introduction - Robotized Transcranial Magnetic Stimulation - page 13

1.3.2 Industrial Robot Design

Already in 2000, Narayana et al. mounted a TMS coil to a neurosurgical robot,
called NeuroMate r (Renishaw plc., New Mills, Gloucestershire, United King-
dom). The NeuroMate r is based on a five joint serial kinematics [ 52 ]. In that
study, they showed the applicability of NeuroMate r in a typical PET/TMS study.
A couple of years later, Lancaster et al. [ 38 ] extended the NeuroMate r robot with
a sixth joint allowing for coil rotations. They evaluated the robot using a head
model and reported a positioning accuracy of roughly 2 mm [ 38 ]. However, as no
tracking system was used for this setup, the patient's head was immobilized.
Matthäus et al., constructed a robotized TMS system which is based on an
industrial robot and a stereo-optic tracking system [ 46 , 47 ]. In this way it combines
the benefits of neuro-navigation and automation. For this approach a common six-
joint industrial robot is adapted to the TMS requirements. While the first approach
focuses on rTMS treatments and standardized setups, this system features high
flexibility and extensibility and additionally lower hardware costs.
Therefore, robotized TMS seems to be a promising and useful tool for TMS
research. Offering a maximum flexibility, all well-established TMS coils and
stimulators can be registered and used with the robotized TMS system with the
industrial robot design. Due to its large workspace and sufficient power reserves,
the system can actively compensate even for spontaneous head movements in the
full robot workspace [ 64 ].
In contrast to the first system, the robot is not equipped with force sensors. It
therefore cannot correct for hair and noise in the head scans using automatic
pressure control. Currently, a manual coil distance adjustment is used for treatment
in which the investigator semi-automatically moves the coil down towards the head
until the subject confirms confirms contact between coil and head. There is no
ongoing pressure adjustment to keep contact between head and coil during stim-
ulation and motion compensation, either. A major drawback of this open system is,
that patient safety and collision-freeness are hard to achieve. Currently, all
potentially critical robot trajectories are forbidden by the control software in order
to achieve collision-freeness. In the remaining configuration space, the robot cannot
approach all possible targets directly and in many cases, the user has to coarsely
pre-position the robot by hand before it can target safely the patient's head.
In this work, the focus lies on this setup by Matthäus to further improve the
system to a safe and clinically applicable robotized TMS system. Therefore, the
current system setup is described in some more detail.

1.3.2.1 Current Setup of the Robotized TMS System

The current setup of the robotized TMS system is shown in Fig. 1.7 . The two main
components are a six-joint industrial robot, in this case an Adept Viper s850
(Adept

Technology, Inc., Pleasanton, CA, USA), and a Polaris stereo-optic