Measuring individual morphological relationship of cortical regions

Abstract

h i g h l i g h t s• A new metric was proposed to quantify individual morphological relations of regions.• The new metric is indexed as the similarity of different morphological distributions.• We estimate the morphological distribution from individual MRI image.• The metric seemed to have potential application to individual differences studies. a r t i c l e i n f o b s t r a c tBackground: Although local features of brain morphology have been widely investigated in neuroscience, the inter-regional relations in brain morphology have rarely been investigated, especially not for individual participants. New method: In this paper, we proposed a novel framework for investigating this relation based on an individual's magnetic resonance imaging (MRI) data. The key idea was to estimate the probability density function (PDF) of local morphological features within a brain region to provide a global description of this region. Then, the inter-regional relations were quantified by calculating the similarity of the PDFs for pairs of regions based on the Kullback-Leibler (KL) divergence. Results: For illustration, we applied this approach to a pre-post intervention study to investigate the longitudinal changes in morphological relations after long-term sleep deprivation. The results suggest the potential application of this new method for studies on individual differences in brain structure.Comparison with existing methods: The current method can be employed to estimate individual morphological relations between regions, which have been largely ignored by previous studies. Conclusions: Our morphological relation metric, as a novel quantitative biomarker, can be used to investigate normal individual variability and even within-individual alterations/abnormalities in brain structure.

Morphological connectivity correlates with trait impulsivity in healthy adults

“…To investigate how the variability of impulsivity relates to the MC of the ACC, we first quantified the MC value between the ACC and each cortical region with the newly-proposed method (Kong et al, 2014) for each participant. Then, we conducted association analyses separately for each hemisphere (that is, only ipsilateral relations were considered) to investigate the possible associations.…”

“…1, the calculation of morphological connectivity (Kong et al, 2014) can be summarized in the following three steps:…”

“…The MC ranges from 0 to 1, where 1 is for two identical distributions. For more technical detail, please see Kong et al (2014).…”

“…However, the covariance method is based on the correlation between subjects and thus can only be used with a large population, and it is incapable of measuring MCs for single subjects, which largely limits its application to individual differences studies in brain structure. To overcome the limitation, here we applied our newly-proposed approach to measuring MC for each individual from the MRI data (Kong et al, 2014). The new approach is based on similarity of regional morphological distributions and has shown its excellent test-retest reliability and validation reflecting human brain architecture and neuroplasticity (Kong et al, 2014; Kong et al, 2015; Wang et al, 2016).…”

Adolescent Tuning of Association Cortex in Human Structural Brain Networks

“…Still, an advantage of structural correlation networks over structural connectomes derived from diffusion imaging using tractography is the relative simplicity of the structural MRI acquisitions compared with diffusion imaging, which in light of its longer acquisition is more prone to motion artefacts (Yendiki et al 2014), and within which tractography presents considerable challenges (Thomas et al 2014; Reveley et al 2015; Maier-Hein et al 2016). Efforts to derive measures of individual contribution to structural correlation networks (Saggar et al 2015) or fully individual networks from structural imaging (Tijms et al 2012; Kong et al 2014, 2015) including through the combination of multimodal features (Seidlitz et al 2017) should increase the practical applicability of structural correlation network research.…”