Biomedical Engineering Reference
Nevertheless, precise coil positioning on the head is not possible with such a
device. Commonly, the TMS coil is coarsely positioned on the head and the patient
moves the head inside the MRI head coil while being stimulated until a clear MEP
can be measured [ 13 ]. The subject's head is then fixated to maintain the posi-
tioning during measurement and image acquisition. Clearly, this coil positioning
technique is not optimal. We therefore propose to develop a robotized TMS system
that can be used for precise coil positioning inside an MRI scanner. For this
ambitious task different major challenges must be solved:
• Design of a manipulator/robot that is MRI-compatible having at least six DOF,
with a workspace that allows to fully operate inside the MRI scanner tube.
• Implementation of a sophisticated tracking method, that works in the magnetic
environment of the MR scanner and is able to track the patient's head within the
limited line of sight due to MRI scanner, head coil and robot.
• Development of trajectory planning and collision avoidance for the limited
volume of the MRI scanner, that also takes the head coil as an additional
obstacle into account.
• Implementation of a method for robot to tracking system calibration. Standard
hand-eye calibration methods might not be applicable due to the specific setup
and space limitations.
With such a robotized setup, systematic research of the functionality of TMS will
be possible. Beside answering research questions also the treatment using rTMS
might profit from such a systematic investigation. For the treatment of chronic
tinnitus, for instance, the stimulation target is commonly located with fMRI. Here,
the subject listens to sound with different frequencies, until the subject confirms
that the tinnitus roughly equals to the current sound frequency. With fMRI the area
of the auditory cortex which is responsible for this sound frequency is detected.
This area is now used as the target for the rTMS treatment stimulation [ 18 ]. Even
though this target localization seems appropriate, only roughly fifty percent of the
patients treated with rTMS profit from the stimulation [ 17 ]. The reason why the
treatment has an effect or not is still unclear [ 12 ]. As 10-15 % of the population
are affected by tinnitus and in 1-2 % the daily life is severely restricted due to this
affection [ 14 ], effective treatment plans are important. With precise coil posi-
tioning inside the scanner, different targets in the auditory cortex or auditory
pathways can be tested by applying a stimulation train and measuring the effect
directly with fMRI. In this way, an optimal stimulation site might be found which
can than be used for rTMS treatment stimulations outside the MRI scanner.
Even though this demands a new robot design, the basic principles of the safe
robotized TMS system can be used as starting point. For instance, the robust
calibration method could be adapted to allow an additional calibration to the MRI
scanner tube and the head coil. Measurement of forces is mandatory for the limited
workspace in the scanner tube, therefore the developed FTA sensor might be
further developed for the MRI environment.