Abstract

Humans developed the ability to use their hands as tools to actively interact with the surrounding world. On a daily basis, without effort or a conscious will to act, we can manipulate dozens of objects placed in the environment to achieve different purposes. This, at first glance simple, ability relies on a network of cerebral areas computing a series of complex neural processes. The core cerebral regions involved in transforming visual inputs into hand actions have been identified within two distinct pareto-frontal networks, the dorsomedial and dorsolateral pathways. These pathways have been traditionally considered to process different types of information independently. The dual-route model of the hand motor system represents a milestone within the literature exploring the neural correlates of motor control. Nevertheless, following the recent improvement of the technical and analytical approaches adopted to investigate brain’s functioning, this initial description has been considered reductive. Nowadays, it is clear that the neural processes needed to produce a hand action require the integration of different information (e.g. spatial location of the object, spatial location of the arm, grip aperture, goal to pursue, etc.) that imply the exchange of information between anatomically distant, but functionally interconnected, cerebral areas. In light of the most recent neuroscientific advance in the field of motor control, we investigated the specific role of the cerebral areas of the motor system and their functional interactions during the planning and execution of hand actions. In the present thesis, we adopted several methodologies (TMS, fMRI) and analysis approaches (univariate, MVPA, DCM), to explore the involvement of defined areas in specific motor tasks, their representational content and their connectivity profiles. In the three neuroimaging studies here presented, we first considered the functional dynamics that occur within the hand network (Chapter 2). Secondly, we described a broader and integrated hand motor system including also the ventral stream, always considered as specialized in perceiving visual inputs (Chapter 3). Finally, we focused on the often neglected homologous regions within the right hemisphere (Chapter 4), providing a pan of the hand motor system in its entirety. The first study (Chapter 2) adopted a combined TMS-fMRI approach, and focused on understanding the interactions between the dorsomedial and dorsolateral pathways of the hand motor system. We adopted a delayed-reach-and-grasp task, performed under different perceptual conditions (eyes opened or closed), and we perturbed the activity of SPOC in the dorsomedial pathway of the left hemisphere by means of rTMS. We used univariate and multivariate analysis to investigate the modifications occurring during the planning phase of the action within areas functionally connected with the region stimulated with TMS. We found that when the normal activity of SPOC is altered, changes in encoding grasping action information occur within the dorsolateral pathway. This study showed a causal interaction between the dorsomedial and dorsolateral pathway of the hand motor network, which are traditionally considered to be specialized and independent. In the second study (Chapter 3), we adopted fMRI to explore the possible communication between dorsal and ventral stream, verifying the possible complementary and supportive role of the temporal cortex in motor control. To this aim, we adopted a delayed tool-pantomiming task known to recruit the ventral stream. Our delayed pantomiming task allowed us to consider the planning phase of the movement together with the execution of the pantomime. With multivariate analysis, we explored where in the dorsal and in the ventral streams different abstract goals of an action, i.e. independent from the tool identity, are represented in respect to more concrete aspects, related to the tool considered in the pantomime. In addition, we investigated the possible functional interactions between temporal and fronto-parietal regions, showing an exchange of information between the two pathways both with MVPA and connectivity analysis (DCM). Overall, these results point out a hand motor system that not only relies on the specialized-for-action dorsal network, but also on temporal lobe areas. In the third study of the thesis (Chapter 4), we combined data from Chapter 3 with a complementary fMRI session. In the second session we changed instruction modality and effector used to perform the pantomime, while experimental design and the task requirement were unchanged. This approach allowed focusing on understanding: (i) changes in the encoding of concrete and abstract representation based on task requirements (i.e. different instruction modality and effector) and (ii) the possible encoding of tool pantomimes’ information also outside the classically-defined left-lateralized tool network, in homologous regions within the right hemisphere. Overall, we found task-dependent changes in the representational content of the considered areas both in the left and in the right hemisphere. These results provided novel insights into the neural correlates of tool pantomime, pointing towards the supportive role of the temporal cortices and of the right hemisphere when planning and pantomiming this type of action. Overall, our studies contributed delineating a novel view on the organization of the hand motor system describing (i) the functional specialization of its different cerebral areas and (ii) the interactions occurring between these regions. Our research highlighted how the hand motor system has a functionally interconnected organization in which, to different degrees, various areas located in three main cerebral routes (ventral stream, dorsolateral and dorsomedial pathways) communicate to build a meaningful motor output.