Functional brain networks in Developmental Topographical Disorientation

Abstract

Despite a decade-long study on Developmental Topographical Disorientation, the underlying mechanism behind this neurological condition remains unknown. This lifelong selective inability in orientation, which causes these individuals to get lost even in familiar surroundings, is present in the absence of any other neurological disorder or acquired brain damage. Herein, we report an analysis of the functional brain network of individuals with Developmental Topographical Disorientation ($n = 19$) compared against that of healthy controls ($n = 21$), all of whom underwent resting-state functional magnetic resonance imaging, to identify if and how their underlying functional brain network is altered. While the established resting-state networks (RSNs) are confirmed in both groups, there is, on average, a greater connectivity and connectivity strength, in addition to increased global and local efficiency in the overall functional network of the Developmental Topographical Disorientation group. In particular, there is an enhanced connectivity between some RSNs facilitated through indirect functional paths. We identify a handful of nodes that encode part of these differences. Overall, our findings provide strong evidence that the brain networks of individuals suffering from Developmental Topographical Disorientation are modified by compensatory mechanisms, which might open the door for new diagnostic tools.

Keywords: Developmental Topographical Disorientation; fMRI; functional connectivity; network neuroscience; resting-state networks.

Fig. 1

(a) Comparison of the connectivity weight distribution in the HC and DTD groups showing that, on average, the DTD group has a greater connectivity and connectivity strength. (b) RSNs in HC and DTD. Fully connected functional connectivities in both groups with their RSN labels extracted from our brain atlas that include ganglia and thalamus (BG-THAL), auditory network and posterior insula (AUDnet-PINS), mesolimbic network (LIMnet), default mode network lateral (DMnet-l), visual network (VISnet), default mode network posteromedial (DMnet-pm), somatomotor network (MOTnet), default mode network anteromedial and left angular gyrus (DMnet-am-lhAG), ventral visual stream and dorsal visual stream (VVIS-DVIS), fronto parietal network (FPnet), and ventral attention network and salience network (VATTnet-SALnet). Size of each RSN, i.e. number of nodes that belongs to each RSN, from top to bottom is 12, 8, 19, 6, 26, 14, 12, 12, 16, 20, and 24, respectively. We reordered the nodes in the adjacency matrices based on their module assignment, as such each block represents the communication within a specific RSN or between a pair of RSNs.

Fig. 2

(a) Characteristic path length of the average functional brain networks in HC and DTD groups as a function of connectivity density. The asterisks indicate significant difference (P < 0.05 using a permutation testing procedure with = 10,000) between the two groups. Note that for denser connectivities, individuals with DTD (filled circles) show significantly smaller values than controls (open circles) over a wide range of thresholds. (b) Average clustering coefficient compared between HC and DTD groups shows a significantly larger in DTD at higher densities. (c) The same plot as in (a) but computed in the case of binary networks showing that the substantial differences in the weighted networks (shown in a) disappear. Specifically, while sometimes the difference in between the two groups remain significant (P < 0.05), they are all of the order of and, hence, tiny. Note that the values in DTD are slightly above those of HC until around 60% connectivity density where they both approach almost the same value of . (d) The same plot as in (b) but computed for binary networks. Most of the significant differences found in the weighted networks (shown in b) disappear in the binary case. The remaining significant differences between of the two groups at 89% density and higher were tiny and close to zero (of the order of or smaller). The significant differences at lower densities (i.e. 71% and smaller) showed on average a difference of around 0.02, yet their P-values were all between 0.01 and 0.05 making them barely significant.

Fig. 3

Using a permutation testing procedure, we identified significant module level differences between the two groups. Highlighted blocks indicate where there is a significant average absolute connectivity difference (P < 0.05, uncorrected) between DTD and HC functional brain networks. These blocks indicate that there is an enhanced connectivity between some RSNs but not within them. The RSN labels are the same as in Fig. 1.

Fig. 4

(a) Comparison of the connectivity weight distribution in HC and DTD networks obtained with partial correlation method (b) Fully connected functional connectivities in both groups with their RSN labels extracted from our brain atlas, which are the same as in Fig. 1. Here, the functional network of each group was inferred through taking pairwise partial correlation of their concatenated BOLD signals. We observed that the majority of strongly connected links, i.e. those with a relatively high partial correlation value, were found in within-module blocks (i.e. diagonal blocks or intra-RSNs). The ventral visual stream and dorsal visual stream (VVIS-DVIS) module is highlighted as the module with the most significant difference between HC and DTD.

Fig. A5

Same as Fig. 1 but using the concatenation method.

Fig. A6

Same as Fig. 3 but using the concatenation method.

Fig. B7

(a, b) Comparison of the percentage of betweenness centrality values across a range of connectivity densities between the DTD and HC groups for two nodes, right ventrolateral visual network (MIST 175) and left dorsolateral fusiform gyrus (MIST 180), where asterisks indicate densities at which the difference between the two groups was statistically significant (P < 0.05). The percentage (or, equivalently, the relative rank) of betweenness centrality values was chosen to understand the actual betweenness centrality values better and in a more qualitative way. Note that, at a certain density, the node (or nodes) with the smallest betweenness centrality value in a given network has a percentage of 100%, whereas the node (or nodes) with the highest betweenness centrality has the smallest percentage, which is % where is the number of nodes in the network. In general, we first rank all values at each density and then multiply the rank by 100 and then divide it by the maximum possible rank at that density. (c, d) As in (a, b) but for the percentage of weighted degrees. The node MIST 180 was particularly selected because of frequently showing significant differences at low connectivity densities.