Segmentation of brain magnetic resonance images based on multi-atlas likelihood fusion: testing using data with a broad range of anatomical and photometric profiles

Abstract

We propose a hierarchical pipeline for skull-stripping and segmentation of anatomical structures of interest from T1-weighted images of the human brain. The pipeline is constructed based on a two-level Bayesian parameter estimation algorithm called multi-atlas likelihood fusion (MALF). In MALF, estimation of the parameter of interest is performed via maximum a posteriori estimation using the expectation-maximization (EM) algorithm. The likelihoods of multiple atlases are fused in the E-step while the optimal estimator, a single maximizer of the fused likelihoods, is then obtained in the M-step. There are two stages in the proposed pipeline; first the input T1-weighted image is automatically skull-stripped via a fast MALF, then internal brain structures of interest are automatically extracted using a regular MALF. We assess the performance of each of the two modules in the pipeline based on two sets of images with markedly different anatomical and photometric contrasts; 3T MPRAGE scans of pediatric subjects with developmental disorders vs. 1.5T SPGR scans of elderly subjects with dementia. Evaluation is performed quantitatively using the Dice overlap as well as qualitatively via visual inspections. As a result, we demonstrate subject-level differences in the performance of the proposed pipeline, which may be accounted for by age, diagnosis, or the imaging parameters (particularly the field strength). For the subcortical and ventricular structures of the two datasets, the hierarchical pipeline is capable of producing automated segmentations with Dice overlaps ranging from 0.8 to 0.964 when compared with the gold standard. Comparisons with other representative segmentation algorithms are presented, relative to which the proposed hierarchical pipeline demonstrates comparative or superior accuracy.

Reduced subcortical volumes among preschool-age girls and boys with ADHD

“…Integral to this estimation is identification of the optimal diffeomorphism acting on the background space of coordinates which affects the evolution with least energy from the randomly selected atlas image to the subject image. The atlas used in this study for subcortical structure segmentation is based on a highly reliable manual parcellation schema and has been demonstrated to produce subcortical segmentations with a high degree of correspondence to the manual ‘gold-standard’ labels (Tang et al, 2015) In addition, scans were visually inspected to ensure that the automated parcellations were properly delineated. The MALF pipeline has been validated in pediatric and elderly clinical populations and has been shown to be superior to other automated algorithms, such as Freesurfer, for performing subcortical segmentation (Tang et al, 2015).…”

“…In addition, we have used an automated subcortical segmentation procedure that has shown to be superior to other segmentation algorithms, such as those implemented in Freesurfer (Tang et al, 2015). Given previous findings of sex differences in anomalous basal ganglia morphology among children with ADHD (Qiu et al, 2009; Seymour et al, 2017), we specifically compared subcortical volumes among girls and boys with ADHD to same-sex TD children.…”

Shape and diffusion tensor imaging based integrative analysis of the hippocampus and the amygdala in Alzheimer's disease

“…For each of the 51 T1-weighted images, volumetric segmentations of the bilateral hippocampi and amygdalas were automatically obtained from a hierarchical segmentation pipeline [44] consisting of two steps; skull-stripping and brain structure segmentation. Built upon a two-level diffeomorphic multi-atlas likelihood-fusion algorithm in the framework of the random deformable template model [45], this segmentation algorithm has demonstrated superior performance when compared to other state-of-the-art methods, especially when applied to subcortical structures including the hippocampus and the amygdala.…”

“…Built upon a two-level diffeomorphic multi-atlas likelihood-fusion algorithm in the framework of the random deformable template model [45], this segmentation algorithm has demonstrated superior performance when compared to other state-of-the-art methods, especially when applied to subcortical structures including the hippocampus and the amygdala. Detailed evaluations of the segmentation accuracy of this method can be found in our previous studies [44–46]. For the current study, we used 16 atlases (T1-weighted images) belonging to the same age range as the participants of this study.…”

“…For inter-group comparisons of the localized shape morphometrics of the bilateral hippocampi and amygdalas, we employed the same linear regression model as described in [20], namely $y_{k}false(sfalse)={\beta }_{k,0}+{\beta }_{k,1}gfalse(sfalse)+\sum _{\text{cov}}{\alpha }_{\text{cov}}{X}_{\text{cov}}false(sfalse)+{\epsilon }_{k}false(sfalse)$, where y k ( s ) is the deformation marker for subject s at vertex k of the structure-specific common template surface, g ( s ) is a binary group variable, X cov ( s ) denotes the covariate information of subject s included in the analysis (in this study, we co-varied for age, gender, as well as the total intracranial volume (TIV) computed from the size of the brain's image after skull-stripping [44,53]), and ε k ( s ) denotes a Gaussian noise structure. We tested the null hypothesis $H_k^0:{\beta }_{k,1}=0$ against the general hypothesis $H_k^1:{\beta }_{k,1}\ne 0$ with the complete null hypothesis being $H_k^0:{\beta }_{k,1}=0$ simultaneously for all k .…”