MSM: a new flexible framework for Multimodal Surface Matching☆

By Emma Claire Robinson, Saâd Jbabdi, Matthew F. Glasser, Jesper L. R. Andersson, Gregory C. Burgess, M. Harms, Stephen M. Smith, David C. Van Essen, Mark Jenkinson. NeuroImage, 2014

665 citations recorded by the source corpus.

Surface-based cortical registration methods that are driven by geometrical features, such as folding, provide sub-optimal alignment of many functional areas due to variable correlation between cortical folding patterns and function. This has led to the proposal of new registration methods using features derived from functional and diffusion imaging. However, as yet there is no consensus over the best set of features for optimal alignment of brain function. In this paper we demonstrate the utility of a new Multimodal Surface Matching (MSM) algorithm capable of driving alignment using a wide variety of descriptors of brain architecture, function and connectivity. The versatility of the framework originates from adapting the discrete Markov Random Field (MRF) registration method to surface alignment. This has the benefit of being unconstrained by choice of a similarity measure and relatively insensitive to local minima. The method offers significant flexibility in the choice of feature set, and we demonstrate the advantages of this by performing registrations using univariate descriptors of surface curvature and myelination, multivariate feature sets derived from resting fMRI, and multimodal descriptors of surface curvature and myelination. We compare the results with two state of the art surface registration methods that use geometric features: FreeSurfer and Spherical Demons. In the future, the MSM technique will allow explorations into the best combinations of features and alignment strategies for inter-subject alignment of cortical functional areas for a wide range of neuroimaging datasets.

Subfields: Cognitive Neuroscience, Pediatrics, Perinatology and Child Health, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/7e8f82daf1d6e7229ad72137d18969db82142ad1

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Pediatrics, Perinatology and Child Health

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

Generative modeling of brain maps with spatial autocorrelation (2020, positive): “To produce the T1w/T2w maps, high resolution T1-and T2-weighted images were first registered to a standard reference space using an areal-feature-based technique (Glasser et al., 2016a; Robinson et al., 2014), then corrected for bias-field intensity inhomogeneities (Glasser and Van Essen, 2011; Glas”
Human brain mapping: A systematic comparison of parcellation methods for the human cerebral cortex (2017, methodology): “, 2012), using cortical surface based alignment (MSMsulc); implemented using Multimodal Surface Matching (Robinson et al., 2014).”
A distributed brain network predicts general intelligence from resting-state human neuroimaging data (2018, methodology): “We use surface-based multi-modally aligned cortical data (MSM-All [95]), together with a parcellation that was derived from this data using an objective semi-automated neuroanatomical peer-reviewed) is the author/funder.”
Spatial Topography of Individual-Specific Cortical Networks Predicts Human Cognition, Personality and Emotion (2017, methodology): “Consequently, despite the use of functionally aligned HCP data, we avoided assuming vertex-level functional correspondence provided by the MSMAll functional alignment (Robinson et al., 2014).”
neuromaps: structural and functional interpretation of brain maps (2022, background): “Transforming individual, subject-level data between different representations and coordinate systems is nontrivial and has been the focus of significant research over the past several decades [5, 15, 30, 54, 63, 64, 71, 89, 93].”
Multimodal Surface Matching with Higher-Order Smoothness Constraints☆ (2017, methodology): “In terms of folding-based methods, both sMSMSTR and SD produced much lower and smoother distortions than FS; although it is important to note both were optimised for alignment of areal features (Robinson et al., 2014; Yeo et al., 2010), whereas FS was optimised for alignment of cortical folds.”
Functional system and areal organization of a highly sampled individual human brain (2015, background): “Registration strategies have been proposed to improve alignment between subjects taking into account functional variability (Robinson et al., 2014; Sabuncu et al., 2010).”
TRIPLE REPRESENTATION OF LANGUAGE, WORKING MEMORY, SOCIAL AND EMOTION PROCESSING IN THE CEREBELLUM: CONVERGENT EVIDENCE FROM TASK AND SEED-BASED RESTING-STATE FMRI ANALYSES IN A SINGLE LARGE COHORT (2018, methodology): “All fMRI data used in the present study included 2mm spatial smoothing and areal-feature aligned data alignment (“MSMAll”, Robinson et al., 2014).”
Development of large-scale functional networks from birth to adulthood: a guide to neuroimaging literature (2017, background): “New multimodal approaches that also anchor surface-based registrations to other structural or functional characteristics of an individual are likely to improve subject to subject registrations, and hence reduce intersubject variability, even more (e.g. (Glasser et al., 2016b; Robinson et al., 2014).”
The Human Connectome Project’s Neuroimaging Approach (2016, methodology): “Areal-feature aligned data were recently released by the HCP (dubbed “MSMAll” because they use the multimodal surface matching algorithm 10 and all of the useful modalities listed above).”
Image processing and Quality Control for the first 10,000 brain imaging datasets from UK Biobank (2017, methodology): “…include surface generation via Freesurfer (Dale et al., 1999), adding morphometric measures such as cortical thickness or cortical surface area, mapping resting fMRI and dMRI onto the cortical surface, and using Multimodal Surface Matching (Robinson et al., 2014) for surface-based registration.”
Local-Global Parcellation of the Human Cerebral Cortex From Intrinsic Functional Connectivity MRI (2017, methodology): “In addition, the HCP utilized a function-based inter-subject registration algorithm (Robinson et al. 2014), which might reduce inter-subject variability and therefore increased connectional homogeneity.”
The impact of traditional neuroimaging methods on the spatial localization of cortical areas (2018, methodology): “MSMAll has previously been shown to provide major improvements in independent task fMRI data that were not used in the registration (20, 21) over MSMSulc.”
Mapping the human brain’s cortical-subcortical functional network organization (2018, positive): “Third, we increased the spatial specificity and anatomical fidelity of the group-level RSFC data by using a high-resolution surface-based analysis in combination with multimodal (functional and anatomical) alignment across subjects (Robinson et al., 2014).”
Building a Science of Individual Differences from fMRI. (2016, background): “A multimodal surface matching (MSM) framework [31] was recently introduced that not only performs similarly to, and faster than, state-of-the-art algorithms based solely on cortical folding [21], but that can also flexibly accept other types of data (e.”
The Developing Human Connectome Project: a Minimal Processing Pipeline for Neonatal Cortical Surface Reconstruction (2017, methodology): “Average thickness maps were calculated by registering the sulcal depth maps of the subjects with MSM (Robinson et al., 2014) to an average template constructed using the method described in Bozek et al.”