Automatic skull segmentation from MR images for realistic volume conductor models of the head: Assessment of the state-of-the-art

Abstract

Anatomically realistic volume conductor models of the human head are important for accurate forward modeling of the electric field during transcranial brain stimulation (TBS), electro- (EEG) and magnetoencephalography (MEG). In particular, the skull compartment exerts a strong influence on the field distribution due to its low conductivity, suggesting the need to represent its geometry accurately. However, automatic skull reconstruction from structural magnetic resonance (MR) images is difficult, as compact bone has a very low signal in magnetic resonance imaging (MRI). Here, we evaluate three methods for skull segmentation, namely FSL BET2, the unified segmentation routine of SPM12 with extended spatial tissue priors, and the skullfinder tool of BrainSuite. To our knowledge, this study is the first to rigorously assess the accuracy of these state-of-the-art tools by comparison with CT-based skull segmentations on a group of ten subjects. We demonstrate several key factors that improve the segmentation quality, including the use of multi-contrast MRI data, the optimization of the MR sequences and the adaptation of the parameters of the segmentation methods. We conclude that FSL and SPM12 achieve better skull segmentations than BrainSuite. The former methods obtain reasonable results for the upper part of the skull when a combination of T1- and T2-weighted images is used as input. The SPM12-based results can be improved slightly further by means of simple morphological operations to fix local defects. In contrast to FSL BET2, the SPM12-based segmentation with extended spatial tissue priors and the BrainSuite-based segmentation provide coarse reconstructions of the vertebrae, enabling the construction of volume conductor models that include the neck. We exemplarily demonstrate that the extended models enable a more accurate estimation of the electric field distribution during transcranial direct current stimulation (tDCS) for montages that involve extraencephalic electrodes. The methods provided by FSL and SPM12 are integrated into pipelines for the automatic generation of realistic head models based on tetrahedral meshes, which are distributed as part of the open-source software package SimNIBS for field calculations for transcranial brain stimulation.

A flexible workflow for simulating transcranial electric stimulation in healthy and lesioned brains

“…9). While our approach for skull segmentation tends to overestimate the skull caudally and 574 occipitally, along the superior sagittal sinus, it slightly underestimates the thickness dorsally where the 576 magnitude (50). However, the general agreement in the change of the magnitude of the electrical field 577 strength indicates that our modeling workflow does not introduce unexpected alterations to the head 578 model.…”

“…The implementation of the outlined workflow by current tES simulation pipelines 48 does not entirely cover use cases with suboptimal imaging data, the presence of pathological tissue in 49 patients or alternative electrode shapes. 50 For instance, MRI data from large-scale imaging studies usually have not been primarily acquired for 51 the purpose of computational head modeling. Performing simulation studies based on such data can 52 become difficult due to challenges in the segmentation of low-contrast tissue such as skull using 53 standard segmentation approaches.…”

“…First, there is no restriction concerning 535 the topology of the sub-compartments of the mesh. As the boundaries are determined directly from a 536 labeled image, there is no requirement of overlap-free boundaries of sub-compartments(50).537Therefore, the inclusion of structures that do not obey a strictly nested arrangement, for example, 538 tumorous or lesioned tissue or holes in the skull(12), is facilitated. Second, image-based meshing is 539 less sensitive to the quality of the input data which avoids extensive postprocessing of the 540 segmentation images.…”

Modeling Primary Fields of TMS Coils with the Fast Multipole Method

“…The dipoles are weighted by the loop current and the subarea sizes. This method has further been used in experimental and theoretical studies (see Nummenmaa et al, 2013;Madsen et al, 2015) and in the well-known, open-source transcranial brain stimulation modeling software SimNIBS Opitz et al, 2015;Nielsen et al, 2018). The method of magnetic dipoles is closely linked to a reciprocity principle (Heller and van Hulsteyn, 1992; Nummenmaa et al, 2013).…”

Software Toolkit for Fast High-Resolution TMS Modeling

“…The chief finite element TMS solver is the well-known, open-source transcranial brain stimulation modeling software SimNIBS Opitz et al, 2015;Nielsen et al, 2018), whose most recent version, v2.1 , currently uses the open-source FEM software getDP. There are also many general-purpose, open-source solvers for finite element modeling, from high-level environments such as getDP (Dular et al, 1988), Deal.II (Bangerth et al, 2007), and FEniCS (Logg et al, 2012), to lower-level environments such as PETSc (Balay et al, 2018).…”