Grant: CIHR Fellowship
Université de Montréal
$90,000.00
How do we know what we know: neuropsychology, neuroimaging, and machine learning unraveling the neuro-cognitive substrate of semantic knowledge.
EF 2019-20
Résumé
Our ability to collect knowledge on objects, facts and people, lies at the core of human intelligence. This uniquely human capacity of storing information about the world is called semantic knowledge. For centuries, these mental representations have been object of scientific investigation: for each concept, a set of features (e.g., "purple, sweet") is linked to an arbitrary symbol (e.g., "berry"), and the information used adaptively to perform the task at hand (e.g., "I will eat it").
In my research, exploiting multivariate methods of data analysis, I demonstrated that different brain areas encode complementary dimensions of concepts: perceptual ones (e.g., it's purple) are stored in posterior, occipital areas, while conceptual ones (e.g., it's a fruit) in anterior, temporal areas. Together with a growing body of literature, these results indicate that semantic knowledge is encoded in distributed yet specialized neural networks. However, open questions remain as to how they are implemented in the brain and what are the earliest sign of their breakdown, denoting neuro-cognitive issues.
Our proposal integrates methods from neuropsychology (e.g., behavioral experiments in both healthy individuals and neurodegenerative patients), cognitive neuroimaging (i.e., functional magnetic resonance imaging and magnetoencephalography), and machine learning to deepen our understanding of the neurobiology of semantic knowledge. We will collect extensive multimodal data which will allow us to better understand the structure of concepts across subjects and tasks, compare cognitive, computational, and neural representations, and test generalization across tasks and stimuli modalities.
By bridging different levels of description, our results will improve current theoretical accounts on semantic knowledge, and pave the way to clinical interventions in patients with impaired semantic processing due to neurodegeneration, a crucial step as we face population aging.
Programme
CIHR Fellowship
Bénéficiaire
Emplacement
Quebec, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
CIHR Funding Decisions Database
Tel que publié
Id
413115
Name
Borghesani Valentina
Region
Québec
Country
Canada
Orgtype
Institution Paid; Research Institution
Instname2
Neurosciences, Mental Health and Addiction
Theme2
Biomedical
Prcname2
Fellowships - Post-PhD
Projecttitle
How do we know what we know: neuropsychology, neuroimaging, and machine learning unraveling the neuro-cognitive substrate of semantic knowledge.
Programtype2
Fellowship Programs
Orgnameinp2
Université de Montréal
Cihramount2
$90,000
Cihrequipment2
$0
Programname2
CIHR Fellowship
Keyworddelim
Aging; Artificial Intelligence; Dementia; Functional Magnetic Resonance Imaging; Language; Machine Learning; Magnetoencefalography; Semantic
Abstract
Our ability to collect knowledge on objects, facts and people, lies at the core of human intelligence. This uniquely human capacity of storing information about the world is called semantic knowledge. For centuries, these mental representations have been object of scientific investigation: for each concept, a set of features (e.g., "purple, sweet") is linked to an arbitrary symbol (e.g., "berry"), and the information used adaptively to perform the task at hand (e.g., "I will eat it").
In my research, exploiting multivariate methods of data analysis, I demonstrated that different brain areas encode complementary dimensions of concepts: perceptual ones (e.g., it's purple) are stored in posterior, occipital areas, while conceptual ones (e.g., it's a fruit) in anterior, temporal areas. Together with a growing body of literature, these results indicate that semantic knowledge is encoded in distributed yet specialized neural networks. However, open questions remain as to how they are implemented in the brain and what are the earliest sign of their breakdown, denoting neuro-cognitive issues.
Our proposal integrates methods from neuropsychology (e.g., behavioral experiments in both healthy individuals and neurodegenerative patients), cognitive neuroimaging (i.e., functional magnetic resonance imaging and magnetoencephalography), and machine learning to deepen our understanding of the neurobiology of semantic knowledge. We will collect extensive multimodal data which will allow us to better understand the structure of concepts across subjects and tasks, compare cognitive, computational, and neural representations, and test generalization across tasks and stimuli modalities.
By bridging different levels of description, our results will improve current theoretical accounts on semantic knowledge, and pave the way to clinical interventions in patients with impaired semantic processing due to neurodegeneration, a crucial step as we face population aging.
Competitiondate
201910
Approvedterm2
2 yrs 0 mth
Pinamesdelim
Borghesani, Valentina
Approvedterm
2 ans 0 mois
Cihrcontribution2
~ $90,000~ $0~