Model-based image updating in deep brain stimulation with assimilation of deep brain sparse data

Abstract

Background: Accuracy of electrode placement for deep brain stimulation (DBS) is critical to achieving desired surgical outcomes and impacts the efficacy of treating neurodegenerative diseases. Intraoperative brain shift degrades the accuracy of surgical navigation based on preoperative images.

Purpose: We extended a model-based image updating scheme to address intraoperative brain shift in DBS surgery and improved its accuracy in deep brain.

Methods: We evaluated 10 patients, retrospectively, who underwent bilateral DBS surgery and classified them into groups of large and small deformation based on a 2 mm subsurface movement threshold and brain shift index of 5%. In each case, sparse brain deformation data were used to estimate whole brain displacements and deform preoperative CT (preCT) to generate updated CT (uCT). Accuracy of uCT was assessed using target registration errors (TREs) at the Anterior Commissure (AC), Posterior Commissure (PC), and four calcification points in the sub-ventricular area by comparing their locations in uCT with their ground truth counterparts in postoperative CT (postCT).

Results: In the large deformation group, TREs were reduced from 2.5 mm in preCT to 1.2 mm in uCT (53% compensation); in the small deformation group, errors were reduced from 1.25 to 0.74 mm (41%). Average reduction of TREs at AC, PC and pineal gland were significant, statistically (p ⩽ 0.01).

Conclusions: With more rigorous validation of model results, this study confirms the feasibility of improving the accuracy of model-based image updating in compensating for intraoperative brain shift during DBS procedures by assimilating deep brain sparse data.

Keywords: biomechanical modeling; deep brain stimulation; image updating.

Figure 1:

The model-based image-updating computational pipeline. *postCT is used in this simulation study for sparse data extraction.

Figure 2:

Illustration of BCs. The fluid plane (green) was perpendicular to the direction of gravity (red) which was rotated 10° from the direction (black) pointing towards center of Earth. Nodes in magenta and red correspond to free nodes, and blue nodes are fixed.

Figure 3:

Sparse data extraction. Displacements were extracted by registering ventricles from preCT to those in postCT using a demons-based diffeomorphic algorithm.

Figure 4:

Deep brain landmark locations at AC, PC, C1 (pineal gland), C2, C3, and C4. Circle center coordinates were used as coordinates of each landmark.

Figure 5:

Surface sparse data (blue) for Case 5 (A) and Case 9 (B) extracted by registering preCT (red) and postCT (green) homologous surface points.

Figure 6:

Target localization errors averaged from pairwise distances among three identifications from each surgeon for all cases.

Figure 7:

Correlation between brain shift at AC, PC and brain shift index.

Figure 8:

3D rendered overlays of homologous ventricle pairs for case 9 (Group L) and case 3 (Group S). From left to right are ventricles in 3D rendered brain masks (A, D), pre- (magenta) and postCT (green) ventricles (B, E), and post- (green) and uCT (blue) ventricles (C, F).

Figure 9:

Qualitative assessment of frontal lobe surface and lateral ventricles for Case 9 (top) and Case 3 (bottom). From left to right are preCT (A, F), uCT (B, G), and postCT (C, H), absolute difference image between preCT and postCT (D, I), and absolute difference image between uCT and postCT (E, J). Cross hairs indicate tips of ventricles in preCT and their corresponding positions after brain shift.

Figure 10:

Illustration of deep brain landmark locations at AC, PC, C1 (pineal gland), C2, C3, and C4 from pre- (red), post- (green), and uCT (blue) in 3D rendered preCT brain masks corresponding to Case 9 (left) and Case 3 (right).

Figure 11:

(A) Comparison of $\mathit{\text{TR}}{E}_{\mathit{\text{preCT}}}$ (red) and $\mathit{\text{TR}}{E}_{\mathit{\text{uCT}}}$ (green) at AC, PC, C1, C2 and C3 for Group L (left) and S (right). (B) Comparison of overall $\mathit{\text{TR}}{E}_{\mathit{\text{preCT}}}$ and $\mathit{\text{TR}}{E}_{\mathit{\text{uCT}}}$ and their three orthogonal components in Group L (large deformation, n=4) and Group S (small deformation, n=6). *p < 0.05, **p < 0.01, ***p < 0.001. Diamonds indicate outliers.