Biomedical Engineering Reference
In-Depth Information
sensor and measure the gravity compensated forces and torques at rest as error
quantification. Second, we apply a realistic setup for coil placement on the head.
Therefore, we ask inexperienced users to operate the optimized hand-assisted
positioning method to accurately place the TMS coil at targets on a human head
phantom. For each target, we measure the error in coil positioning. In contrast to
an evaluation during a real TMS application, we employ a human head phantom to
accurately measure the positioning error within a realistic scenario.
7.4.1 Coil Calibration and Gravity Compensation
To estimate the application-oriented accuracy of the coil calibration, we mount
seven different TMS coils to the FT-controlled robotized TMS system. The
parameters of the coils are listed in Table 5.1 . We first perform an optimized coil
calibration for each coil and use the calibrated values for gravity compensation of
the FTA sensor. Subsequently, we rotate the robot randomly and record the gravity
compensated forces and torques. For each coil we record approximately
8,000-10,000 data points. Now, we estimate the error of the gravity compensation
as the difference of the recorded compensated forces and torques to zero. In this
way, we also estimate the maximum error of the optimized coil calibration as it is
the basis for gravity compensation. Note that the MC-B35 figure-of-eight coil is
not available anymore and cannot be used for this evaluation.
7.4.2 Precision of Optimized Hand-Assisted Positioning
We evaluate the precision of coil positioning with the optimized hand-assisted
positioning method as a correlate for its effectivity, efficiency and usability. To this
end, we have prepared a Maxstim 70 mm circular coil such that we have rigidly
inserted a felt tip pen into the coil's center as illustrated in Fig. 7.5 a. In this way,
the pen is orientated perpendicular to the coil surface. Subsequently, we attach the
coil to the coil holder and perform a coil calibration for the FTA sensor.
For evaluation we ask ten inexperienced users to perform hand-assisted posi-
tioning. For familiarization with the system, we ask the operators to position the
coil with the pen as the focus point at three targets with different coil rotation
angles as displayed in Fig. 7.5 b.
As the main task we use a realistic scenario. We therefore mark six distinct
targets on a head phantom made of styrofoam as shown in Fig. 7.6 . Each target
coarsely corresponds to targets that are frequently used for TMS:
• The Primary Motor Leg Area (M1-LEG) is of interest for brain research
applications (cf. Chap. 3 ).
 
Search WWH ::




Custom Search