Neurocognitive networks mediating cognitive flexibility: Implications for neuropsychiatric disorders

点击翻译按钮获取中文摘要

Individual variability has predominantly been investigated in the context of reward-based learning. Yet, another aspect of human flexibility crucial for adaptive behaviour includes structural/statistical learning i.e. learning temporal sequences without explicit reward (Kourtzi, & Welchman, 2019). When performing a structural learning task, healthy individuals have been shown to adapt to changes in the environment's temporal statistics -unbeknownst to them - and use probability matching and maximization decision strategies (Wang et al., 2017; Karlaftis et al., 2019). Interestingly, there is growing body of evidence to suggest that the brain architecture supporting learning predictive structures without explicit reward involves cortico-striatal circuits that have been previously implicated in reward-based learning (Robbins, 2007). Specifically, adopting a strategy closer to matching related to plasticity in motor corticostriatal circuits, while selecting the most probable outcomes (maximization) related to plasticity in visual, motivational and executive corticostriatal circuits (Giorgio et al., 2018; Karlaftis et al., 2019). Further analysis is required to better understand the shared brain mechanisms mediating learning with and without reward contingencies. In addition, different decision strategies may be more efficacious in different contexts. A matching strategy has been associated with exploration and may be the optimal strategy when the environment changes (Iigaya et al.,2019) whereas maximization has been associated with exploitation and maybe the optimal strategy when under risk or uncertainty (Schulze, van Ravenzwaaij, & Newell, 2015). The proposed research seeks to determine whether a structural learning training paradigm could be utilised to optimize adaptive behaviour. Deficits in cognitive flexibility occur in OCD and many possibly related mental health disorders (e.g., addiction, Tourette Syndrome and eating disorders). OCD may be regarded as a prototypical disorder of compulsivity and an example of a "behavioural addiction" (Robbins & Clark, 2015). Therefore, the study of the neurocognitive networks underpinning cognitive flexibility in this disease model has broader relevance for the field of psychiatry. It is posited here that a structural learning intervention could be used for the symptomatic treatment of OCD. Aims:(1) The proposed research aims to examine the construct of cognitive flexibility, initially in a healthy population. In particular, a large-scale web-based assessment of cognitive flexibility and structural learning will be administered to general-population samples. (2) To identify sub-processes of cognitive flexibility and subsequently link these processes to specific functional and anatomical networks using multimodal imaging techniques. Further, the inter-relations between structural learning and reversal learning will be examined by use of computational methods.(3) To test the effects of a structural learning intervention on cognitive flexibility in individuals with and without OCD. Methods and techniques:(1) This research will compare a number of tests of cognitive flexibility and structural learning in healthy individuals. From these measures, latent variables of cognitive flexibility will be constructed and sub-processes that are conceptually and empirically separable will be identified by conducting a factor analysis. It is hypothesized here that there will be two differential aspects to cognitive flexibility relating to exploratory and exploitative strategies. (2) To map neurocognitive networks mediating the different aspects of cognitive flexibility the imaging protocol with consist of a combination of neuroimaging techniques such as functional magnetic resonance imaging, and magnetic resonance spectroscopy. Combined DTI-based segmentation analysis and rs-fMRI of the striatum will be used to compute the functional and structural connectivity between the striatal