Proposal Description

Understanding the complexities of the human brain requires advanced analytical techniques that can capture its dynamic and continuous activity patterns. Traditional methods in neuroscience often focus on static snapshots of brain activity, overlooking the rich temporal and spatial dynamics inherent in neural processes. Functional Data Analysis (FDA), which has become one of the most prolific fields of statistics since the mid-80s, offers a compelling solution to this challenge by providing tools to analyze brain data as smooth functions. Recent advancements in neuroimaging and physiological recording techniques, like fMRI and EEG, have significantly enhanced our ability to investigate brain function. However, analyzing neuroscientific data poses numerous challenges, including artifact contamination, manifold representation, and various issues associated with the curse of dimensionality. One key reason to apply FDA methods in neuroscience is its ability to uncover the temporal and spatial dynamics of brain function and to enable researchers to characterize individual differences, which is essential for understanding the heterogeneity of neurological and psychiatric disorders.

The objective of this IPS is to showcase the potential of new advances in FDA for modeling brain data, thereby advancing our understanding of brain function to improve diagnosis and treatment of neurological diseases. More specifically, the papers presented in the session will concern with the following topics:

- Shape analysis of anatomical structures using neuroimaging data.
- Integrative analysis of functional and high-dimensional data. Application to the Human Connectome Project.
- Outlier detection in fMRI task experiments through multivariate functional data techniques.
- Multivariate functional independent component analysis for modelling turbulent flows on the cortical field.

The results presented are interested not only for researchers in this topic but also for statisticians and neuroscientists that could apply them in their professional development. All researchers participating in this IPS section are experts in the topic with an excellent research career. The session is organized by Ana M. Aguilera (Full Professor in Statistics at the Department of Statistics and Operation Research at University of Granada) and Marc Vidal (pre-doctoral researcher in Institute for Psychoacoustics and Electronic Music at Ghent University, and Institute of Mathematics at University of Granada).

The chair, Ana M. Aguilera, is a very experienced researcher on FDA and data-driven statistics in different areas such as the environment, spectrometry, biomechanics, neurology and electronics.

Anuj Srivastava is Full professor in the Department of Statistics at Florida State University. He is a highly reputable researcher and his main interest lies in the use of geometry and statistics in advancing inferences related to complex objects.

Ana Arribas-Gil is Visiting Professor in the Department of Statistics at University Carlos III of Madrid. She is specialized in Functional Data Analysis and Longitudinal studies in Molecular Biology.

Marc Vidal is a young researcher with relevant contributions on Functional Independent Component approaches for the blind separation and classification of signals in Neuromusicology. His doctoral thesis will be defended in June 2024 under the co-supervision of Prof. Marc Leman (Director Institute for Psychoacoustics and Electronic Music, Ghent University) and Prof. Ana M. Aguilera.