Automatic segmentation of deep grey nuclei using a high‐resolution 7T magnetic resonance imaging atlas—Quantification of T1 values in healthy volunteers

Abstract

We present a new consensus atlas of deep grey nuclei obtained by shape‐based averaging of manual segmentation of two experienced neuroradiologists and optimized from 7T MP2RAGE images acquired at (.6 mm)3 in 60 healthy subjects. A group‐wise normalization method was used to build a high‐contrast and high‐resolution T1‐weighted brain template (.5 mm)3 using data from 30 out of the 60 controls. Delineation of 24 deep grey nuclei per hemisphere, including the claustrum and 12 thalamic nuclei, was then performed by two expert neuroradiologists and reviewed by a third neuroradiologist according to tissue contrast and external references based on the Morel atlas. Corresponding deep grey matter structures were also extracted from the Morel and CIT168 atlases. The data‐derived, Morel and CIT168 atlases were all applied at the individual level using non‐linear registration to fit the subject reference and to extract absolute mean quantitative T1 values derived from the 3D‐MP2RAGE volumes, after correction for residual B1+ biases. Three metrics (the Dice and the volumetric similarity coefficients and a novel Hausdorff distance) were used to estimate the inter‐rater agreement of manual MRI segmentation and inter‐atlas variability, and these metrics were measured to quantify biases due to image registration, and their impact on the measurements of the quantitative T1 values was highlighted. This represents a fully automated segmentation process permitting the extraction of unbiased normative T1 values in a population of young healthy controls as a reference for characterizing subtle structural alterations of deep grey nuclei relevant to a range of neurological diseases.

Multi‐scale structural alterations of the thalamus and basal ganglia in focal epilepsy using 7T MRI

“…Precise delineation and quantification of thalamic nuclei and basal ganglia have become feasible with the advent of 7T MRI and the possibility to map T 1 both at a submillimeter resolution and (relatively) fast using the MP2RAGE sequence (Marques et al, 2010). As a result, multiple protocols have been proposed, including the one developed in‐house (Brun et al, 2022), to segment thalamic nuclei and basal ganglia from such data (Datta et al, 2021; Su et al, 2019). When we compared results from our in‐house 7TAMIBrain atlas method with those from another popular approach (THOMAS), thalamic volumetric measurements appear relatively similar despite differences in segmentation algorithm and nuclei definitions.…”

Multi-scale structural alterations of the thalamus and basal ganglia in focal epilepsy as demonstrated by 7T MRI

“…Precise delineation and quantification of thalamic nuclei and other basal ganglia have become feasible with the advent of 7T MRI and the possibility to map T 1 both at a submillimeter resolution and (relatively) fast using the MP2RAGE sequence 26 . As a result, multiple protocols have been proposed, including the one developed in-house 32 , to segment thalamic nuclei and basal ganglia from such data 33,38 . When we compared results from our in-house 7TAMIBrain atlas method with those from another popular approach (THOMAS), thalamic volumetric measurements appear relatively similar despite differences in segmentation algorithm and nuclei definitions.…”

“…Isolation of brain structures was achieved in two-fold: (i) through an in-house developed thalamic nuclei (and basal ganglia) atlas ('7TAMIBrain') 32 and (ii) by applying the Thalamus Optimized Multi Atlas Segmentation ('THOMAS') method 33 . Both methods rely on multi-atlas label fusion (MALF) which has proven superior compared to single-atlas registration methods 34 .…”

“…An improved version of the recently published 7TAMIBrain pipeline by Brun et al was employed 32 . While initially based on a single template-to-subject registration, current version makes use of 60 atlases constructed from the 7TAMIBrain dataset 35 .…”

Future Prospects of Positron Emission Tomography–Magnetic Resonance Imaging Hybrid Systems and Applications in Psychiatric Disorders

“…Depending on the region of interest (ROI), different MRI modalities can be selected. For imaging the subnuclei of the thalamus, T1-weighted MRI with inversion recovery appears to be more suitable for maximizing the contrast of the subnuclei by nullifying the signal of a specific nucleus [ 14 , 15 , 16 , 17 ]. T2*-weighted imaging with a gradient echo pulse sequence can achieve better contrast between the white and gray matter.…”