Academic journal article Cognitie, Creier, Comportament

What Can Neuroimaging Tell Us about the Early Development of Visual Categories?

Academic journal article Cognitie, Creier, Comportament

What Can Neuroimaging Tell Us about the Early Development of Visual Categories?

Article excerpt


While brain imaging studies of visual cognition have contributed extensively to our understanding of the different mechanisms involved in object processing and categorization, in adulthood, infancy studies have only started to employ these techniques. We identify in this paper a few of the methodological and theoretical reasons that hindered a more enthusiastic use of imaging methods. Focusing on three theoretical questions that stand out from the infant object categorization literature we show that, when the methodology is adapted to the study of young populations and the interpretations guided by equivalent results from the adult literature, brain imaging can help shed light on cognitive development.

KEYWORDS: infants, categorization, brain imaging, ERPs, language.

Brain imaging methods have greatly contributed to the understanding of visual cognition, in adulthood. Guided by the results obtained with more invasive methods in monkeys, human brain imaging techniques, that vary both in temporal and spatial resolution, have provided access to different dimensions of visual processing. These techniques fall into two categories, those measuring the brain electrical activity - the Electroencephalography (EEG) and the Magnetoencephalography (MEG) and those measuring the brain activity indirectly, by detecting changes in the cerebral hemodynamics - Functional Magnetic Resonance (fMRI) and recently the Near Infrared Spetroscopy (NIRS). fMRI has greatly helped to understand both the anatomical and functional organization of the visual areas in the brain. EEG and MEG have allowed us to chart the temporal unfolding of object processing in real time. More recently, with the advent of better localization algorithms, these electrophysiological methods have also begun to provide good spatial information about the location of neural processes, albeit with lower resolution than fMRI.

By using well-established experimental paradigms in conjunction with the above imaging techniques, it is now possible to explore a variety of brain related cognitive issues such as the selectivity of a type of process and the developmental time-course of this specialization, the effect of expertise on neural processes, and the modularity or interactivity of different types of cognitively relevant neural processes. Such questions are of major relevance for understanding cognitive development. A very rich collection of behavioural data has shown that infants' perception of their visual environment changes dramatically from the first minutes of life until adulthood. We believe that understanding these changes can be greatly helped by our knowledge of brain development.

Despite this great potential, the number of infant neuro-imaging studies is still relatively small. Numerous ethical and methodological barriers largely prevent fMRI and MEG from being employed with young healthy children. Both these methods are very sensitive to motion artefacts and thus require the immobilization of the infant's head. fMRI also involves subjecting the baby to radio frequency magnetic pulses and to high levels of noise. However, where procedural solutions that minimize infants' discomfort during MRI scanning have been found1, valuable information was obtained in diverse domains such as the localization of languagespecific areas in infants as young as three months of age (Dehaene-Lambertz, Dehaene, & Hertz-Pannier, 2002) and the connectivity between different subcortical structures and cortical areas (Dubois, Hertz-Pannier, Dehaene-Lambertz, Cointepas, & Le Bihan, 2006). Moreover, with very young participants, there is the problem of trying to limit the frequent motion artefacts. However, with these disadvantages also comes a big advantage, these imaging methods do not require any explicit behavioural response from the participant. This is good news for studying early development because it sidesteps the problem of what behavioural methods to use. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.