How brain networks tic: Predicting tic severity through rs-fMRI dynamics in Tourette syndrome
Shukti Ramkiran, Tanja Veselinović, Jürgen Dammers, Arnim Johannes Gaebler, Ravichandran Rajkumar, N. Jon Shah und Irene Neuner
26. May 2023
Tourette syndrome (TS) is a neurological and psychiatric disorder characterised by involuntary movements and vocalisations known as tics. Various theories have been proposed to explain TS, and the recognition of TS as a network problem rather than as isolated disturbances in specific brain regions is growing. In light of this perspective, this study examines the predictive capability of direct and indirect dynamic network metrics in determining tic severity in order to gain a comprehensive understanding of tic pathophysiology.
Functional connectivity analysis was performed on resting-state functional magnetic resonance imaging (fMRI) data acquired from 36 participants using three techniques: static, sliding window dynamic, and an ICA-based estimated dynamic method. This was followed by an examination of the static and dynamic network topo- logical properties. Key factors were identified using a regression model with LASSO regularization, and the validity of the model was assessed using the leave-one-out (LOO) technique.
Thus, this network-based analysis approach helps in understanding how information is processed (both dynamically and overall) by each specific region engaged in the network in relation to tics in TS. It is interesting to note that, while several studies have shown a direct implication of the amygdala in TS, this study demonstrates that the network properties of the amygdala itself are not direct predictors of tic severity; rather, the regions it communicates within the social processing context are. This finding broadens our understanding of the roles played by other regions in the network in addition to the amygdala itself.
The results of the research further indicate dysfunction in the primary motor cortex, the prefrontal-basal ganglia loop, and the amygdala-mediated visual social processing network as being significant in TS and provides new insights into understanding the underlying causes of tics in TS.
This figure below visualises the identified predictors of tic severity. The legend lists the regions, followed by Kendall's correlation between the predictor and tic severity in brackets, followed by the network property identified and its model weight in the second brackets. The identified network properties were the temporal correlation coefficient and temporal average path length of the slow direct dynamic network (sliding window dynamic network) represented by dSW-tCC and dSW-tAPL respectively.
Future work will focus on investigating the network properties of the identified regions in response to treatment, at the individual level. This will help in devising targeted treatment and therapeutic strategies for patients.
Origional publication: How brain networks tic: Predicting tic severity through rs-fMRI dynamics in Tourette syndrome