M1 or M2

Integrative Neuroscience and Cognition Center


INCC - Université de Paris & CNRS
45 rue des Saints-Pères
75006 Paris

Length of internship
6 months
English and/or French
Our phenomenological experience of the visual world is stable and constant – a remarkable fact, given that the input to the visual brain is ever-changing. Indeed, visual input is constantly interrupted by blinks; the positions of objects on the retina shift with every eye movement; the low-level visual features of objects like color and shading change with differences in lighting. Despite this high variability, we do not perceive the world to move every time we move our eyes, nor objects to change color every time the sun moves behind a cloud. The cognitive and cerebral underpinnings of this constructed stability are an area of intense interest in cognitive neuroscience. One mechanism that may bridge the gap between variable input and stable visual experience is if the content of current perception is not (only) the result of immediate sensory information, but rather the result of spatio-temporal integration. Such integration would allow the visual system to capitalize on the fact that the environment is, in the short-term, highly auto-correlated. Evidence for spatio-temporal integration in visual perception comes from a behavioral phenomenon called serial dependence (SD): when asked to report a particular visual attribute, healthy human adult observers are biased towards objects seen in the recent past (e.g. Fischer & Whitney, 2014). The cerebral underpinnings of SD, on which subjective visual stability may depend, are poorly understood, and have not been much studied. The goal of the current project is to study the cerebral underpinnings of SD by examining the phenomenon in brain-lesioned patients, in particular, homonymous hemianopia (HH), a phenomenon caused by retrochiasmal lesions of the visual system (Chokron, Perez & Peyrin, 2016). By comparing the performance of HH patients with lesions of left and right V1 to those of a group of control subjects in multiple SD tasks, we aim to answer three questions: 1) Does SD occur within visual brain areas, and in multiple brain areas? 2) Is residual (unconscious) visual processing in the blind hemifield integrated into subsequent (conscious) perception? 3) Are there hemispheric differences in terms of spatio-temporal integration in visual perception?