Skip to main content
European Commission logo print header

Attention Warps Early Sensory Maps

Final Report Summary - AWESOME (Attention Warps Early Sensory Maps)

Our sensory systems register an overwhelming amount of information from multiple sources. How and where in the brain this is integrated and selected to guide behavior remains an open question. The established view assigns these tasks to the so-called associative areas, whereas primary sensory areas provide a complete and “objective” representation of the input. In the visual system, early visual cortex (EVC, areas V1 and V2) is believed to encode the simple features of the scene, such as light intensity variations, their orientation and their position. Successive stages of processing integrate selected elements of these representations with other sources of information: the state of subject (her body position, eye movements, current behavioral task), signals from other senses, from memory, etc.
The present project was aimed to test a radically different perspective: that the selection and integration of sensory signals happens since the earliest processing stages, so that nowhere in the brain may an “objective” representation of the visual input be found. We used a combination of studies on adult human subjects, testing their behavioral responses to visual stimuli, measuring physiological responses like eye movements and pupillary reflexes, as well as studying the physiology of the visual system by means of neuroimaging techniques (fMRI). The results of these experiments provided strong support to our hypotheses.
We first investigated the perception of spatial position for briefly presented visual stimuli. Early visual areas entail an explicit representation of space: a “retinotopic map”, thought to depend on the anatomical connections from the eye to the cortex and therefore assumed to be stable and immutable, at least in the adult individual. However, we found that maps of perceived visual space are extremely flexible. They change dramatically depending on the behavioral context of the visual stimulation, producing illusory displacements for stimuli flashed while the subject is planning an eye movement or a shift of attention. Importantly, dynamic changes consistent with the perceptual biases accompanying shifts of attention were observed in the maps of responses in V1 (which we measured with fMRI taking advantage of the specialized neuroimaging tools available at the extra-European host). This directly suggests that early retinotopic maps are not stable, but dynamically remap inputs depending on their behavioral context. Additional behavioral experiments revealed a strong parallel between spatial distortions induced by gaze shifts and attention shifts, supporting the view that the two are intimately linked.
Further experiments could take advantage of the most advanced neuroimaging facilities – ultra-high resolution fMRI available at the European Host. With its enhanced spatial resolution, this allows us to investigate the physiology of a small internal structure: the thalamus, specifically one of its nuclei involved in visual processing (called LGN). This is the first brain structure that receives information from the eyes (the retina), which it relays to the primary visual cortex. Preliminary data suggest that, already at this level, the representation of the simplest visual feature (light intensity) is affected by the state of the subject. Visual stimulation normally increases the neural activity in the visual cortex and the thalamus, more so when the stimulation is more intense. However, our preliminary observations suggest that the opposite can happen when the subject’s state is manipulated.
Thus, we have data suggesting that the simplest visual attributes (intensity, position) are represented in a “subjective” way (i.e. in a way that is not only dependent on the stimulus that falls on the eye, but modulated by the subject’s behavior) even at the earliest stages of brain processing. A third line of experiments suggests that these modulatory effects affect even the earliest possible level: the eye. The eye is an optical system with movable parts, like a camera. One of these parts is the pupil, which works as an aperture affecting the amount of light that enters the visual system. It was long believed that the diameter of the pupil is automatically regulated by the amount of ambient light, through a reflex (automatic behavior preserved in unconscious states). However, we found that the regulation of pupil size is affected by the subject’s behavioral context: the pupillary response to light is enhanced or suppressed when the subject’s attention is focused upon or away from the light source. Two independent groups have now replicated this surprising finding.
In summary, the results of our experiments complement and expand in important ways the most recent literature, by showing that the integration and selection of sensory signals and other sources of information is not confined to high level associative areas; it is a fundamental characteristic of all sensory stages, all the way from the cortex to the eye. These observations have a strong impact on the field of neurophysiology at large. Not only do they challenge the view of the brain as a hierarchy of processes of increasing complexity; they also pose the need for new models of the mechanisms mediating the combination and selection of sensory signals, which must be simple and fast enough to interact with the earliest stages of processing.
The implications of our findings, however, go well beyond our field of research. Demonstrating that no separation exists between the acquisition of sensations and their interpretation may have deep philosophical implications – especially in the field of epistemology. Important consequences are also hoped for in medical research and the development of new clinical approaches. Our research suggests that deficits in behavioral and attentional control will affect the processing of sensory information, correctly predicting that important perceptual impairments will occur even in conditions where the sensory system is relatively spared (e.g. Autism Spectrum, ADHD, conditions associated to pre-term birth), and possibly stimulating the development of new perspectives on the rehabilitation and treatment of these conditions.
In conclusion, the AWESoMe research project has had a clearly positive outcome, as measured by three main indicators: (1) an exceptional level of scientific productivity, in terms of number and quality of publications, importance of the results and their impact; (2) the advancement of the researcher’s career, who acquired proficiency in a new set of techniques (neuroimaging), developed original and independent lines of research and thereby formed an international network of scientific contacts, and (3) the establishment of a strong and productive link between the European host laboratory (Pisa, Italy) and one of the most renowned extra-European research centers for neuroimaging (Seattle, WA, U.S.A).
For further information regarding this research, please contact:
The researcher: Dr. Paola Binda, p.binda1@in.cnr.it
The scientist in charge: Prof. Concetta Morrone, concetta@in.cnr.it
Department of Translational Research on New Technologies in Medicine and Surgery, University of Pisa, via San Zeno 31 Pisa, Italy