Determining the orientation angle of directional leads for deep brain stimulation using computed tomography and digital x-ray imaging: A phantom study

Purpose: Orientating the angle of directional leads for deep brain stimulation (DBS) in an axial plane introduces a new degree of freedom that is indicated by embedded anisotropic directional markers. Our aim was to develop algorithms to determine lead orientation angles from computed tomography (CT) and stereotactic x-ray imaging using standard clinical protocols, and subsequently assess the accuracy of both methods. Methods: In CT the anisotropic marker artifact was taken as a signature of the lead orientation angle and analyzed using discrete Fourier transform of circular intensity profiles. The orientation angle was determined from phase angles at a frequency 2/360(circle) and corrected for aberrations at oblique leads. In x-ray imaging, frontal and lateral images were registered to stereotactic space and sub-images containing directional markers were extracted. These images were compared with projection images of an identically located virtual marker at different orientation angles. A similarity index was calculated and used to determine the lead orientation angle. Both methods were tested using epoxy phantoms containing directional leads (Cartesia(TM), Boston Scientific, Marlborough, USA) with known orientation. Anthropomorphic phantoms were used to compare both methods for DBS cases. Results: Mean deviation between CT and x-ray was 1.5(circle) +/- 3.6(circle) (range: - 2.3(circle) to 7.9(circle)) for epoxy phantoms and 3.6(circle) +/- 7.1(circle) (range: - 5.6(circle) to 14.6(circle)) for anthropomorphic phantoms. After correction for imperfections in the epoxy phantoms, the mean deviation from ground truth was 0.0(circle) +/- 5.0(circle) (range: - 12(circle) to 14(circle)) for x-ray. For CT the results depended on the polar angle of the lead in the scanner. Mean deviation was - 0.3(circle) +/- 1.9(circle) (range: - 4.6(circle) to 6.6(circle)) or 1.6(circle) +/- 8.9(circle) (range: - 23(circle) to 34(circle)) for polar angles <= 40(circle) or > 40(circle). Conclusions: The results show that both imaging modalities can be used to determine lead orientation angles with high accuracy. CT is superior to x-ray imaging, but oblique leads (polar angle > 40(circle)) show limited precision due to the current design of the directional marker. (C)2017 American Association of Physicists in Medicine