Conclusion

Results shown in diff may be summarized as follows.

• The distance between the corners found from the images can differ in some cases from the actual distance of as much as 6 mm, meaning that the displacement of individual corners can be around 3 mm. However, the smallest difference found is 0.1 mm. The mean absolute differences are 2.5 mm, 3.0 mm and 3.0 mm for the transverse, sagittal and coronal acquisitions respectively.
• The distance between the corners found from the images can vary by a maximum of 3 mm depending of the type of acquisition. It can also be observed that the distance between the corners is similar for the sagittal and the coronal acquisitions. This can be understood by the fact that these data were acquired with the readout gradient in the same direction. The small difference can however be attributed to eddy current effects, since the gradient switching sequence is different for the two acquisitions, and possibly to differences in linearity between the two phase encoding gradients.

This experiment shows the order of magnitude of the distortion affecting the 3-D structure formed by the markers of the OBT stereotactic frame. According to [20], the largest part of these distortions is caused by the bulk magnetic susceptibility of the frame plates. This demonstrates that special care must be taken in the design (shape and material) of any external localizing device.

It is important to underline the fact that this result does not imply that the error on the localization of points in conventional stereotaxy using the OBT frame is in the order of 5 mm. We have to keep in mind that this experiment investigated the accuracy of the representation of the stereotactic frame as a 3-D structure, whereas conventional stereotaxy localization only relies on information found on one slice.