Geometric effects of volume-to-surface mapping of fMRI data

A cortical surface template for human neuroscience

Neuroimaging data analysis relies on normalization to standard anatomical templates to resolve macroanatomical differences across brains. Existing human cortical surface templates sample locations unevenly because of distortions introduced by inflation of the folded cortex into a standard shape. Here we present the onavg template, which affords uniform sampling of the cortex. We created the onavg template based on openly available high-quality structural scans of 1,031 brains-25 times more than existing cortical templates. We optimized the vertex locations based on cortical anatomy, achieving an even distribution. We observed consistently higher multivariate pattern classification accuracies and representational geometry inter-participant correlations based on onavg than on other templates, and onavg only needs three-quarters as much data to achieve the same performance compared with other templates. The optimized sampling also reduces CPU time across algorithms by 1.3-22.4% due to less variation in the number of vertices in each searchlight.

Spurious correlations in surface-based functional brain imaging

The study of functional MRI data is increasingly performed after mapping from volumetric voxels to surface vertices. Processing pipelines commonly used to achieve this mapping produce meshes with uneven vertex spacing, with closer neighbours in sulci compared to gyri. Consequently, correlations between the fMRI time series of neighbouring sulcal vertices are stronger than expected. However, the causes, extent, and impacts of this bias are not well understood or widely appreciated. We explain the origins of these biases, and using in-silico models of fMRI data, illustrate how they lead to spurious results. The bias leads to leakage of anatomical cortical folding information into fMRI time series. We show that many common analyses can be affected by this "gyral bias", including test-retest reliability, fingerprinting, functional parcellations, regional homogeneity, and brain-behaviour associations. Finally, we provide recommendations to avoid or remedy this spatial bias.

Feature similarity gradients detect alterations in the neonatal cortex associated with preterm birth

The early life environment programmes cortical architecture and cognition across the life course. A measure of cortical organisation that integrates information from multimodal MRI and is unbound by arbitrary parcellations has proven elusive, which hampers efforts to uncover the perinatal origins of cortical health. Here, we use the Vogt-Bailey index to provide a fine-grained description of regional homogeneities and sharp variations in cortical microstructure based on feature gradients, and we investigate the impact of being born preterm on cortical development at term-equivalent age. Compared with term-born controls, preterm infants have a homogeneous microstructure in temporal and occipital lobes, and the medial parietal, cingulate, and frontal cortices, compared with term infants. These observations replicated across two independent datasets and were robust to differences that remain in the data after matching samples and alignment of processing and quality control strategies. We conclude that cortical microstructural architecture is altered in preterm infants in a spatially distributed rather than localised fashion.

Local gradient analysis of human brain function using the Vogt-Bailey Index

In this work, we take a closer look at the Vogt-Bailey (VB) index, proposed in Bajada et al. (NeuroImage 221:117140, 2020) as a tool for studying local functional homogeneity in the human cortex. We interpret the VB index in terms of the minimum ratio cut, a scaled cut-set weight that indicates whether a network can easily be disconnected into two parts having a comparable number of nodes. In our case, the nodes of the network consist of a brain vertex/voxel and its neighbours, and a given edge is weighted according to the affinity of the nodes it connects (as reflected by the modified Pearson correlation between their fMRI time series). Consequently, the minimum ratio cut quantifies the degree of small-scale similarity in brain activity: the greater the similarity, the 'heavier' the edges and the more difficult it is to disconnect the network, hence the higher the value of the minimum ratio cut. We compare the performance of the VB index with that of the Regional Homogeneity (ReHo) algorithm, commonly used to assess whether voxels in close proximity have synchronised fMRI signals, and find that the VB index is uniquely placed to detect sharp changes in the (local) functional organization of the human cortex.

Connectopic mapping techniques do not reflect functional gradients in the brain

Functional gradients, in which response properties change gradually across a brain region, have been proposed as a key organising principle of the brain. Recent studies using both resting-state and natural viewing paradigms have indicated that these gradients may be reconstructed from functional connectivity patterns via "connectopic mapping" analyses. However, local connectivity patterns may be confounded by spatial autocorrelations artificially introduced during data analysis, for instance by spatial smoothing or interpolation between coordinate spaces. Here, we investigate whether such confounds can produce illusory connectopic gradients. We generated datasets comprising random white noise in subjects' functional volume spaces, then optionally applied spatial smoothing and/or interpolated the data to a different volume or surface space. Both smoothing and interpolation induced spatial autocorrelations sufficient for connectopic mapping to produce both volume- and surface-based local gradients in numerous brain regions. Furthermore, these gradients appeared highly similar to those obtained from real natural viewing data, although gradients generated from real and random data were statistically different in certain scenarios. We also reconstructed global gradients across the whole-brain - while these appeared less susceptible to artificial spatial autocorrelations, the ability to reproduce previously reported gradients was closely linked to specific features of the analysis pipeline. These results indicate that previously reported gradients identified by connectopic mapping techniques may be confounded by artificial spatial autocorrelations introduced during the analysis, and in some cases may reproduce poorly across different analysis pipelines. These findings imply that connectopic gradients need to be interpreted with caution.