GraSP: Geodesic Graph-based Segmentation with Shape Priors for the functional parcellation of the cortex

Abstract

Resting-state functional MRI is a powerful technique for mapping the functional organization of the human brain. However, for many types of connectivity analysis, high-resolution voxelwise analyses are computationally infeasible and dimensionality reduction is typically used to limit the number of network nodes. Most commonly, network nodes are defined using standard anatomic atlases that do not align well with functional neuroanatomy or regions of interest covering a small portion of the cortex. Data-driven parcellation methods seek to overcome such limitations, but existing approaches are highly dependent on initialization procedures and produce spatially fragmented parcels or overly isotropic parcels that are unlikely to be biologically grounded. In this paper, we propose a novel graph-based parcellation method that relies on a discrete Markov Random Field framework. The spatial connectedness of the parcels is explicitly enforced by shape priors. The shape of the parcels is adapted to underlying data through the use of functional geodesic distances. Our method is initialization-free and rapidly segments the cortex in a single optimization. The performance of the method was assessed using a large developmental cohort of more than 850 subjects. Compared to two prevalent parcellation methods, our approach provides superior reproducibility for a similar data fit. Furthermore, compared to other methods, it avoids incoherent parcels. Finally, the method’s utility is demonstrated through its ability to detect strong brain developmental effects that are only weakly observed using other methods.

Studying cyto and myeloarchitecture of the human cortex at ultra-high field with quantitative imaging: R1, R2* and magnetic susceptibility

“…To better emphasize the differences between different brain regions a myelo-structure based atlas could have been more appropriate (Nieuwenhuys, 2013) as it is known that within one Brodmann region, several myelin based regions can exist (Nieuwenhuys et al, 2014), the use of Broadman areas as ROI will tend to average out such differences. Another segmentation approach would have been automatic segmentation of the subject specific clusters for different quantitative contrasts similar to the functional parcellation of the cortex (Honnorat et al, 2015). The richness of the cortical information attained in these maps opens the door to more advanced co-registration procedures between subjects as well as its segmentation (Dinse et al, 2015).…”

GraMPa: Graph-Based Multi-modal Parcellation of the Cortex Using Fusion Moves

“…The lower performance for 50 parcels could indicate that it is difficult to obtain large smooth parcels while agreeing with all reliability maps. Our parcellations' agreement with the underlying structure is evaluated by i) computing the average functional coherence (FC) [6] and ii) evaluating the agreement with task fMRI activation maps (obtained using FSL's standard tools) using the Bayesian Information Criterion (BIC) [16]. FC evaluates the average correlation between a parcel's average timeseries and the timeseries of all vertices in the same parcel.…”

“…Similarly to dMRI, we obtain a parcel centre by maximising the within parcel timeseries correlation. Due to the low SNR of fMRI data, we follow the approach proposed in [6] and compute a representative average timeseries for each label. We average the timeseries of the N closest vertices (in terms of shortest path) to the parcel centre that are in the same parcel.…”

“…This kind of approach however assumes a strong reliability of dMRI data and a global agreement between structural and functional connectivity. Markov Random Field (MRF) models have been applied successfully to dMRI and fMRI driven cortical parcellation tasks [6,13]. Their main advantage is their versatility, in the sense that no restriction is made on the data term driving the parcellation scheme.…”

Localization And Quantification Of Intramuscular Damage Using Statistical Parametric Mapping And Skeletal Muscle Parcellation

“…Such a method allows analysis to be performed in a stereotactic space taking into account inter-individual morphology variability. Furthermore, specific parcellations have been used to discriminate several heart 33,34 and brain [35][36][37] areas and could be utilized to quantify the extent of local damage in individuals.…”