Improved Vim targeting for focused ultrasound ablation treatment of essential tremor: A probabilistic and patient-specific approach

Abstract

Magnetic resonance-guided focused ultrasound (MRgFUS) ablation of the ventral intermediate (Vim) thalamic nucleus is an incisionless treatment for essential tremor (ET). The standard initial targeting method uses an approximate, atlas-based stereotactic approach. We developed a new patient-specific targeting method to identify an individual's Vim and the optimal MRgFUS target region therein for suppression of tremor. In this retrospective study of 14 ET patients treated with MRgFUS, we investigated the ability of WMnMPRAGE, a highly sensitive and robust sequence for imaging gray matter-white matter contrast, to identify the Vim, FUS ablation, and a clinically efficacious region within the Vim in individual patients. We found that WMnMPRAGE can directly visualize the Vim in ET patients, segmenting this nucleus using manual or automated segmentation capabilities developed by our group. WMnMPRAGE also delineated the ablation's core and penumbra, and showed that all patients' ablation cores lay primarily within their Vim segmentations. We found no significant correlations between standard ablation features (e.g., ablation volume, Vim-ablation overlap) and 1-month post-treatment clinical outcome. We then defined a group-based probabilistic target, which was nonlinearly warped to individual brains; this target was located within the Vim for all patients. The overlaps between this target and patient ablation cores correlated significantly with 1-month clinical outcome (r = -.57, p = .03), in contrast to the standard target (r = -.23, p = .44). We conclude that WMnMPRAGE is a highly sensitive sequence for segmenting Vim and ablation boundaries in individual patients, allowing us to find a novel tremor-associated center within Vim and potentially improving MRgFUS treatment for ET.

Keywords: 7T MRI; MR guided focused ultrasound; Vim; essential tremor; surgical targeting; thalamus; white matter nulled MPRAGE.

FIGURE 1

Comparison of WMnMPRAGE images with standard T2‐weighted and CSFnMPRAGE images for three representative patients. Coronal, axial, and sagittal slices are shown centered on the Vim nucleus. The greater sensitivity of WMnMPRAGE imaging provides higher contrast between thalamic nuclei than the standard imaging protocols, which allows manual and automated segmentations of these nuclei. Note: Due to the cytoarchitectonically gradual progression between the Vim and the more dorsal part of VLp, the VLp was manually segmented and the Vim (orange arrow) was defined as the ventral half of the VLp. AV, anteroventral; CM, centromedian; Hb, habenula; MD, mediodorsal; MGN, medial geniculate nucleus; MTT, mammillothalamic tract; Pf, parafascicular; Pul, pulvinar; VA, ventral anterior; VLa, ventral lateral anterior; Vim, ventral intermediate; VLp, ventral lateral posterior; VPL, ventral posterior lateral

FIGURE 2

Overview of preprocessing and registration pipeline. Generally, 7T pretreatment WMnMPRAGE images were used for thalamic segmentations; these were linearly registered to the 3T pretreatment WMnMPRAGE volume for each patient, and nonlinear warps were applied to transform them into the group normalized brain template space. 3T post‐treatment images were used to detect and characterize ablation features; these were first linearly registered to the 3T pretreatment WMnMPRAGE volumes and then nonlinearly transformed to the group normalized brain template space. The study‐specific group normalized brain template was created from all subjects' 3T pretreatment WMnMPRAGE images. With all relevant image volumes transformed to the template space, and relevant features of the initial targets and final ablations derived in the template space, we obtained interpatient ablation metrics and conducted correlation analyses with clinical outcome across patients. See Section 2 for further details

FIGURE 3

Study‐specific, group normalized brain template. A study‐specific template was created to enable direct comparison of the Vim and ablation regions across patients. The template was created from the 3T pretreatment WMnMPRAGE scans of 13 patients (scans flipped right to left if ablations were made in the right hemisphere). Image volumes were first nonlinearly registered using ANTs and then averaged to form the normalized average brain. See Section 2 for further details

FIGURE 4

Agreement between manual and THOMAS automated segmentations of the Vim in WMnMPRAGE images. Overlay of manual (filled nuclei) and THOMAS (gray outlines) segmentations in three cases with the best (Patient A, Dice = 0.90), average (Patient C, Dice = 0.74), and worst (Patient M, Dice = 0.55) Dice values for the Vim. Group mean Vim Dice = 0.67 ± 0.07 (see Table 2). Coronal, axial, and sagittal slices are shown centered on the Vim. This analysis used 7T WMnMPRAGE images for all patients, except Patient C, for whom a 3T WMnMPRAGE image was used due to a lack of a 7T WMnMPRAGE image. Thalamic nuclei abbreviations defined in Figure 1

FIGURE 5

Comparison of immediate post‐treatment 3T WMnMPRAGE images and standard T2‐weighted images in three representative patients. Coronal, axial, and sagittal slices are shown centered on the FUS ablation. Note the improved visibility of the ablation in the WMnMPRAGE images compared to the T2‐weighted images, allowing clearer delineation of an outer penumbra (yellow outline), more hyperintense core (cyan outline), and, in many cases, a hypointense region within the core. For all patients, the ablations were located within and around the Vim (orange outline) defined by manual segmentation in the WMnMPRAGE images. Thalamic nuclei abbreviations defined in Figure 1

FIGURE 6

A probabilistic initial target region identified within the Vim. A probabilistic target region (black outline) was defined as the region of 100% overlap of patient whole ablations (blue heatmap), eroded to the mean volume of patient ablation cores. Coronal, axial, and sagittal slices are shown centered on the probabilistic target region in (a) group normalized space with the group thalamic segmentation overlaid and in (b–d) patient native space, by nonlinearly warping the target region from group normalized to patient native space, for three representative patients with their respective manual thalamic segmentations overlaid. Thalamic nuclei abbreviations defined in Figure 1

FIGURE 7

The probabilistic target region identified within the Vim is significantly associated with 1‐month clinical outcome, whereas the standard stereotactically defined target region is not. (a) There is a significant correlation between the fraction of the probabilistic target region covered by the patient's ablation core and 1‐month clinical outcome (CRST A + B), indicating that the greater the overlap between ablation core and probabilistic target region, the better the 1‐month clinical outcome. (b) No significant correlation was observed between the fraction of the standard stereotactic target region covered by the patient's ablation core and 1‐month clinical outcome (CRST A + B)