Imaging patients pre and post deep brain stimulation: Localization of the electrodes and their targets
Yan Li,Sagar Buch,Naying He,Chencheng Zhang,Yingying Zhang,Tao Wang,Dianyou Li,Ewart Mark Haacke,Fuhua Yan
DOI: https://doi.org/10.1016/j.mri.2020.09.016
IF: 3.13
2021-01-01
Magnetic Resonance Imaging
Abstract:<h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Purpose</h3><p>Deep brain stimulation (DBS) has become a widely performed surgical procedure for patients with medically refractory movement disorders and mental disorders. It is clinically important to set up a MRI protocol to map the brain targets and electrodes of the patients before and after DBS and to understand the imaging artifacts caused by the electrodes.</p><h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Methods</h3><p>Five patients with DBS electrodes implanted in the habenula (Hb), fourteen patients with globus pallidus internus (GPi) targeted DBS, three pre-DBS patients and seven healthy controls were included in the study. The MRI protocol consisted of magnetization prepared rapid acquisition gradient echo T1 (MPRAGE T1W), 3D multi-echo gradient recalled echo (ME-GRE) and 2D fast spin echo T2 (FSE T2W) sequences to map the brain targets and electrodes of the patients. Phantom experiments were also run to determine both the artifacts and the susceptibility of the electrodes. Signal to noise ratio (SNR) on T1W, T2W and GRE datasets were measured. The visibility of the brain structures was scored according to the Rose criterion. A detailed analysis of the characteristics of the electrodes in all three sequence types was performed to confirm the reliability of the postoperative MRI approach. In order to understand the signal behavior, we also simulated the corresponding magnitude data using the same imaging parameters as in the phantom sequences.</p><h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Results</h3><p>The mean ± inter-subject variability of the SNRs, across the subjects for T1W, T2W, and GRE datasets were 20.1 ± 8.1, 14.9 ± 3.2, and 43.0 ± 7.6, respectively. High resolution MPRAGE T1W and FSE T2W data both showed excellent contrast for the habenula and were complementary to each other. The mean visibility of the habenula in the 25 cases for the MPRAGE T1W data was 5.28 ± 1.11; and the mean visibility in the 20 cases for the FSE T2W data was 5.78 ± 1.30. Quantitative susceptibility mapping (QSM), reconstructed from the ME-GRE sequence, provided sufficient contrast to distinguish the substructures of the globus pallidus. The susceptibilities of the GPi and globus pallidus externa (GPe) were 0.087 ± 0.013 ppm and 0.115 ± 0.015 ppm, respectively. FSE T2W sequences provided the best image quality with smallest image blooming of stimulator leads compared to MPRAGE T1W images and GRE sequence images, the measured diameters of electrodes were 1.91 ± 0.22, 2.77 ± 0.22, and 2.72 ± 0.20 mm, respectively. High resolution, high bandwidth and short TE (TE = 2.6 ms) GRE helped constrain the artifacts to the area of the electrodes and the dipole effect seen in the GRE filtered phase data provided an effective mean to locate the end of the DBS lead.</p><h3 class="u-h4 u-margin-m-top u-margin-xs-bottom">Conclusion</h3><p>The imaging protocol consisting of MPRAGE T1W, FSE T2W and ME-GRE sequences provided excellent pre- and post-operative visualization of the brain targets and electrodes for patients undergoing DBS treatment. Although the artifacts around the electrodes can be severe, sometimes these same artifacts can be useful in identifying their location.</p>
radiology, nuclear medicine & medical imaging