Biomedical Engineering Reference
Variations of the different calibration methods for different regions in the robot
Translational variation (mm)
Rotational variation ( )
the tracking system is not as severely affected by possible calibration errors of the
tracking system as the detection of the marker attached to the robot's end effector.
Furthermore, the angle the additional marker is seen at by the tracking system is
126.96.36.199 Accuracy of Coil Positioning—Overall System Error
The most interesting and meaningful evaluation for robotized TMS is the impact
of the utilized calibration method on the system's overall positioning accuracy.
Therefore, we have measured the relationship between the employed calibration
method and the accuracy of coil targeting. To this end, we have selected five
distinct targets on a human head phantom and have measured the positioning error
after coil placement with the robotized TMS system (cf. Sect. 188.8.131.52 ).
We have found that the mean differences from target point to actual coil
position are 1.80 mm for the QR24 algorithm, 7.12 mm for the method by Tsai
and Lenz, and 2.21 mm for the online calibration approach. The single recordings
for each target point can be found in Table 4.4 . The directions of the divergences
to the target are visualized in Fig. 4.12 .
This realistic evaluation of the robotized TMS system's overall positioning
accuracy supports that the presented online calibration method is suitable for
robotized TMS and for accurate coil placement. Coil positioning is only slightly
more accurate (roughly 0.4 mm) when using the QR24 algorithm for hand-eye
calibration instead of the online calibration method. The method by Tsai and Lenz,
however, results in a less accurate coil positioning.
4.4 Benefits for Robotized TMS
We have presented a new method for performing the calibration between robot and
tracking system in a robust online fashion which uses a marker attached to the
robot's third link. Our experimental results have shown that this calibration is