Cognitive Neuroscience Research Group

karlpetersson

Location

Building 1, Laboratory 2.53

Faculdade de Ciências Humanas e Sociais

University of Algarve

Phone: +351 289 800 900

kmpetersson@ualg.pt

Team Members

Karl Magnus Petersson (Coordinator, Principal Investigator)

Alexandra Isabel Dias Reis

Luís Miguel Faísca

Susana Silva

Susana Araujo

Inês Bramão

Ana Teresa Martins

Andreia Pacheco

Dina Silva

Filomena Café Inácio (Phd Student)

About the CNR

The Cognitive Neuroscience (CNS) Research Group is an internationally renowned unit with its research focus on questions related to the functional organization of the human brain, cognition, and behaviour. The CNS group addresses scientific issues related to language processing, reading & writing, developmental dyslexia, object recognition and naming as well as implicit learning. Currently, most group members are senior researchers who integrate cross-disciplinary expertise in neuroscience, cognitive & computational neuroscience, functional neuroimaging, cognitive and experimental psychology.

CNR Objectives

The CNS group has set-up and manages state-of-the-art laboratories for electrophysiological EEG/ERP research, behavioural and eye-tracking research (including a stationary high-resolution platform and a mobile platform), as well as data-analysis and computing facilities. Currently, on-going scientific projects relate to the neural basis of mental processes and their behavioural manifestations, using state-of-the-art methods from experimental cognitive psychology, eye-tracking, and functional neuroimaging (EEG/MRI/TMS). The CNS group has also initiated some work combining methods from behavioural genetics and functional neuroimaging.

Team Members

karl_team

Research Lines

neuro_3
neuro_8
Brain Awareness Week

Language processing and implicit learning

The ability to generate and comprehend language is grounded in the faculty of language, a specific brain system, and the capacity to communicate complex information and thoughts through language, is a result of the way our brains acquire its native language, largely in an implicit manner. The CNS is currently investigating the interaction between natural language syntax and semantics with FMRI in order to understand the neural correlates of how the brain generates sentence-level understanding. A second ongoing objective is to investigate implicit artificial language/grammar learning (AGL), which has re-vitalized the study of language processing, language acquisition, and language evolution during the last decade. We have developed a unique AGL paradigm based on the structural mere-exposure effect in combination with preference/grammaticality classification which we aim to investigate with eye-tracking, FMRI, and EEG methodology.

Developmental dyslexia

Undiagnosed or untreated children with reading & writing disorders are at high risk for academic underachievement, social-emotional problems associated with chronic school failure, underemployment and behavioural problems in adulthood. Recently, we have obtained results suggesting that there is more to developmental dyslexia than a phonological processing deficit and that visual integration impairments might contribute to poor object naming performance. Moreover, some evidence suggests that implicit sequence learning might operates at sub-normal levels in developmental dyslexia. In this research line, the objective is to provide a deeper understanding, at the brain and cognitive levels, of developmental dyslexia by means of eye-tracking, EEG, and FMRI. A second objective is to investigate the word recognition potential (N170) by means of EEG and to establish a direct link between reading strategy and eye-movements by means of eye-tracking.

Neurocomputational models of language processing

The major objective is to develop neurobiologically realistic computational models that can account for core aspects of natural language processing. A central notion is state-dependent processing in adaptive dynamical systems as this applies to recurrent, sparsely connected spiking networks with local adaptive mechanisms at various time-scales. The first goal is to develop a neurobiologically plausible model that functions as a transducer that maps structured sequences onto semantic representations. This system will be implemented in a recurrent spiking network that simulates the interaction between Broca’s and Wernicke’s regions. A second aim is to develop interfaces for the generation of local field potentials, which will further constrain the architecture and the bridge modeling and experimental levels. The third aim is to develop theoretical, computational, and modeling competences in order to take timely advantage of tools, methods, and data that will be made available by large-scale neuroscience projects, including the Human Brain Project, the BRAIN Initiative, the Brain Activity Map and the Allen Institute for Brain Science projects. Long-term it is expected that a deeper understanding of the nature of neural computation will yield new models of language processing that eventually will replace more traditional ones.