Grant: CIHR Fellowship
McGill University
$45,000.00
Make no mistake: acquisition of novel sensorimotor maps without error correction
EF 2016-17
Résumé
As we learn to speak or play a musical instrument, we constantly make errors but because goals and targets change from movement to movement, we may not have the opportunity to go back and correct our mistakes. However, current understanding of motor learning is based almost exclusively on adaptation studies in which learning is based on error correction, and therefore it is unclear how motor learning proceeds when error correction is not possible. The present work hypothesises that early stages of motor learning are not based on error. We designed a novel paradigm as a window into this early stage of learning, in which participants make arm movements to auditory targets and hence have to create a new mapping between movements and associated sounds in order to complete the task. The sounds are not localised in space but, like speech sounds, can only be distinguished by their frequencies. Pilot data reveals that participants can learn this challenging task. Using this task, for the first time we are able to monitor participants as they establish an audio-motor map. We will use computational, behavioural and neuroimaging tools to test the hypothesis that this type of early motor learning is not based on errors. Specifically, (1) we will show with computational simulations that instead of error correction, learning could be driven by a process of collecting a history of movements and their associated sensory effects into a mapping; (2) we then test the simulation predictions behaviourally by presenting a new target on every trial and therefore precluding error correction, and (3) use neuroimaging and neuro-stimulation to show brain areas involved in (somatosensory) memory formation subserve map learning instead of areas involved in error correction. Our work will shed light on early motor learning process not based on error, which may allow therapies capitalise these learning mechanisms for patients who cannot apply error correction because of brain injury.
Programme
CIHR Fellowship
Bénéficiaire
Emplacement
Quebec, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
CIHR Funding Decisions Database
Tel que publié
Id
362125
Name
van Vugt Floris T
Region
Québec
Country
Canada
Orgtype
Institution Paid; Research Institution
Instname2
Neurosciences, Mental Health and Addiction
Theme2
Biomedical
Prcname2
Fellowships - Post-PhD
Projecttitle
Make no mistake: acquisition of novel sensorimotor maps without error correction
Programtype2
Fellowship Programs
Orgnameinp2
McGill University
Cihramount2
$45,000
Cihrequipment2
$0
Programname2
CIHR Fellowship
Keyworddelim
Action-Perception Coupling; Computational Modeling; Ctbs; Feedback Error-Based Learning; Fmri; Motor Learning; Sensorimotor Integration; Sensorimotor Maps
Abstract
As we learn to speak or play a musical instrument, we constantly make errors but because goals and targets change from movement to movement, we may not have the opportunity to go back and correct our mistakes. However, current understanding of motor learning is based almost exclusively on adaptation studies in which learning is based on error correction, and therefore it is unclear how motor learning proceeds when error correction is not possible. The present work hypothesises that early stages of motor learning are not based on error. We designed a novel paradigm as a window into this early stage of learning, in which participants make arm movements to auditory targets and hence have to create a new mapping between movements and associated sounds in order to complete the task. The sounds are not localised in space but, like speech sounds, can only be distinguished by their frequencies. Pilot data reveals that participants can learn this challenging task. Using this task, for the first time we are able to monitor participants as they establish an audio-motor map. We will use computational, behavioural and neuroimaging tools to test the hypothesis that this type of early motor learning is not based on errors. Specifically, (1) we will show with computational simulations that instead of error correction, learning could be driven by a process of collecting a history of movements and their associated sensory effects into a mapping; (2) we then test the simulation predictions behaviourally by presenting a new target on every trial and therefore precluding error correction, and (3) use neuroimaging and neuro-stimulation to show brain areas involved in (somatosensory) memory formation subserve map learning instead of areas involved in error correction. Our work will shed light on early motor learning process not based on error, which may allow therapies capitalise these learning mechanisms for patients who cannot apply error correction because of brain injury.
Competitiondate
201611
Approvedterm2
1 yr 0 mth
Pinamesdelim
van Vugt, Floris T
Approvedterm
1 an 0 mois
Cihrcontribution2
~ $45,000~ $0~