Traditional approaches to the study of large-scale human brain activity rely on the tools of network science, which conventionally model neural dynamics as a network of pairwise interactions between brain regions. However, these methods are limited in their focus on dyadic interactions, which are insufficient to capture the brain's complex functional organization.
In this Thesis, we leveraged a novel framework based on topological data analysis to address this limitation. Specifically, we applied this framework in the context of functional magnetic resonance imaging (fMRI), to infer higher-order interactions—interactions involving more than two brain regions simultaneously—from time series of nodal activity.

The novel contribution of this work lies in the application of the mentioned higher-order framework to the domain of emotional research. In particular, we employed a multimodal dataset designed to explore the interplay between emotional experience and neural activity during the viewing of short films.
Our findings reveal that, unlike traditional lower-order approaches, higher-order methods effectively identify temporal clusters of brain activity aligned with film-elicited emotional patterns, and distinguish between emotional states of opposite valence.
Overall, the higher-order topological approach offered novel insights into the complex organization of brain group dynamics underlying emotional experiences, paving the way for future research directions.