Optimization and comparative evaluation of nonlinear deformation algorithms for atlas-based segmentation of DBS target nuclei

Abstract

Nonlinear registration of individual brain MRI scans to standard brain templates is common practice in neuroimaging and multiple registration algorithms have been developed and refined over the last 20 years. However, little has been done to quantitatively compare the available algorithms and much of that work has exclusively focused on cortical structures given their importance in the fMRI literature. In contrast, for clinical applications such as functional neurosurgery and deep brain stimulation (DBS), proper alignment of subcortical structures between template and individual space is important. This allows for atlas-based segmentations of anatomical DBS targets such as the subthalamic nucleus (STN) and internal pallidum (GPi). Here, we systematically evaluated the performance of six modern and established algorithms on subcortical normalization and segmentation results by calculating over 11,000 nonlinear warps in over 100 subjects. For each algorithm, we evaluated its performance using T1- or T2-weighted acquisitions alone or a combination of T1-, T2- and PD-weighted acquisitions in parallel. Furthermore, we present optimized parameters for the best performing algorithms. We tested each algorithm on two datasets, a state-of-theart MRI cohort of young subjects and a cohort of subjects age- and MR-quality-matched to a typical DBS Parkinson’s Disease cohort. Our final pipeline is able to segment DBS targets with precision comparable to manual expert segmentations in both cohorts. Although the present study focuses on the two prominent DBS targets, STN and GPi, these methods may extend to other small subcortical structures like thalamic nuclei or the nucleus accumbens.

Toward a unified connectomic target for deep brain stimulation in obsessive-compulsive disorder

“…To minimize this issue, we used a modern neuroimaging pipeline that has been scientifically validated in numerous studies and involved advanced concepts such as brain shift correction 83 , multispectral normalization, subcortical refinement 83 and phantom-validated electrode localizations 84 . The normalization strategy that was applied was found to automatically segment the STN as precisely as manual expert segmentations 85 and each step of the pipeline was carefully assessed and corrected if needed by a team with long-standing expertise in this area 86,87 . Besides, both post-operative CT (thirty-three patients) and post-operative MRI (seven-teen patients) were used for electrode localization in the current dataset.…”

A framework for evaluating correspondence between brain images using anatomical fiducials

“…The focal template misregistrations we have identified in fMRIPrep with AFIDs are meant to serve as a baseline for refinement in future versions that can be compared transparently and potentially incorporated for testing new versions as part of a continuous integration workflow. Using additional image contrasts (Xiao et al, 2017) or subcortical tissue priors (Ewert et al, 2019) to drive template registration have been demonstrated using conventional voxel overlap techniques to result in more optimal registrations that can also be tested using the AFIDs framework.…”

Left prefrontal impact links subthalamic stimulation with depressive symptoms

“…In some patients where this strategy failed, a multispectral implementation of the Unified Segmentation approach implemented in Statistical Parametric Mapping software (SPM12;http://www.fil.ion.ucl.ac.uk/spm) was applied. These two methods are available as pre-sets in Lead-DBS and were top-performers to segment the STN with precision comparable to manual expert segmentations in a recent comparative study 22 . DBS contacts were automatically pre-reconstructed using PaCER or the TRAC/CORE approach and manually refined if needed 20 .…”