Conscious – unconscious dissociations in visual perception: clues from hemianopic patients

CONSCIOUS – UNCONSCIOUS DISSOCIATIONS IN VISUAL PERCEPTION: CLUES FROM HEMIANOPIC PATIENTS. Lesions along the visual pathway produce characteristic gaps in the visual field. The most common type of homonymous field disorder (HVFD) is hemianopia. The probability of occurrence of HVFDs following a stroke, trauma or surgery is high, around 20-30%. Patients with HVFDs are seriously impaired in everyday activities and this is because of the acquired disability in visual orienting and exploration. Recovery of visual function is possible but is rarely complete. However, in the seventies of last century it has been shown by Poeppel et al. and subsequently by Weiskrantz et al. that some hemianopic patients retain some form of visually guided behaviour in the blind hemifield in absence of perceptual awareness (“blindsight”). This Thesis is mainly based on casting new light on this intriguing phenomenon. In particular, the purpose of the present research project is multifold: i) To investigate whether higher-order perceptual effects like Numerosity detection and Gestalt phenomena are present in the blind hemifield of hemianopic patients; ii) To investigate the basic psychophysical properties of conscious as well as unconscious (blindsight) residual vision in the blind field and iii) To study the anatomy and functionality of the intact and damaged hemispheres through electrophysiological recordings and functional imaging. In the first project, we employed a redundant signal effect (RSE) paradigm, in healthy participants and in patients, to investigate the perceptual characteristics of the intact and blind hemifield, and the interaction between them. Initially, we tested in healthy participants whether the RSE increases with 4 vs 1 stimuli. We used two different configurations of stimuli, in one the four redundant stimuli formed a gestalt-like figure in the shape of a diamond, while in the other the position of stimuli was randomized. We tested 18 participants (mean age: 31.1). They were required to keep their fixation steady on a central fixation point and, following onset of an acoustic warning signal, to press a button as quickly as possible upon presentation of the visual stimuli. Results showed an increase of the RSE (i.e. faster reaction times) with four with respect to single stimuli independently from the configuration proposed. The analysis of the present data showed that the RSE effect in the Single bilateral presentation and in the Quadruple Gestalt bilateral presentation is attributable to neural coactivation rather than probability summation, in that there was a violation of Miller’s race inequality. Conversely, the result obtained in the Quadruple Random bilateral presentation was attributable to a probability summation, in that there was no violation of the race inequality. In the second part of this project, we investigated if this procedure was more likely to increase the probability to detect blindsight in hemianopic patients and in hemisperectomized patients. We tested six patients with hemianopia (mean age: 45.83) as a result of cortical or optic tract lesion and one hemisperectomized patient. The procedure and design were the same as in the previous experiment. The results showed a trend toward an increase of speed of reaction time as measured by the RSE with multiple stimuli presented simultaneously to the intact and blind hemifield in comparison to stimuli to the intact field only. It is interesting to point out that the RSE in hemianopics and in the hemisperectomized patient was more pronounced for the gestalt-like configuration and this suggests the presence of gestalt-like perception in blindsight and provides important clues for rehabilitation. Moreover the functional magnetic resonance imaging results showed as expected an activation of visual areas localized at the level of the ventral visual pathway. In the second project, we tried to understand which of the two hemispheres mediates visually guided behavior following stimulus presentation to the blind hemifield of patients with hemianopia. We tested 9 patients (mean age:44.5) with the Poffenberger paradigm (PP): a RT task that involves a lateralized visual stimulus presentation in the intact and in the blind hemifield and a unimanual response. The crossed-uncrossed difference (CUD), a behavioural estimate of callosal interhemispheric transfer (IT) time for stimuli presented to the intact hemifield of all patients was positive and this can be interpreted as related to the lesioned hemisphere subserving the response whilst in the blind hemifield the results were more variable with some patients showing a positive and others a negative CUD (i.e. the response was subserved by the intact hemisphere) and this could be related to the extension of the individual lesions. In the second part, we used a new method for testing blindsight, namely to compare the cumulative distribution frequency (CDF) of RT for stimulus presentation to the intact and the blind field under normal or occluded viewing conditions. We tested 6 patients and the results were that two of them showed blindsight. We also used functional magnetic resonance (fMRI) to assess the involvement of specific brain structures in the IT of visuomotor information and the role of the intact hemisphere. We tested with the Poffenberger paradigm the well known blindsight patient GY. Behavioural results showed a positive CUD for the stimulation of the intact hemifield, and a negative CUD for the stimulation of the blind hemifield, suggesting that the intact hemisphere mediates the blindsight responses. Functional results showed activation of cortical areas (ventral visual areas, dorsal visual areas, sensori-motor and premotor areas) and callosal areas, with interesting differences in the conscious and unconscious conditions (anterior vs. middle-posterior, respectively). In the third project, we measured the electrophysiological response to visual stimulation of blind and intact visual field in the Poffenberger Paradigm (PP) in 12 healthy subjects and in three hemianopic patients. The purpose of the investigation was to inquire into the nature and the functionality of the lesioned hemisphere as well as try to understand the efficiency of IT. In both hemispheres we found an absence of ERP responses for stimulation of the hemianopic field while for stimulation of the intact hemifield we found a good direct response in the contralateral hemisphere. Interestingly, we observed a different response in the three patients for the indirect commissural response: the response recorded in the lesioned hemisphere was absent or very weak in posterior sites for patient EA and patient PC while for patient SL the response recorded in the lesioned hemisphere was comparable to that of healthy participants. This suggests that IT and integration of visual information from the intact to the damaged hemisphere of hemianopic patients seems to occur at the perceptual stage depending on the functional and anatomical integrity of the splenium of the CC. In short, the thrust of this entire Thesis is multifold. First, by using the Redundant Signal Effect it was found that there is a substantial cross-talk between the two hemispheres in hemianopic patients, including one hemispherectomy patient, and that unconscious vision is sensitive to stimulus numerosity and shows a sort of gestaltic perceptual organization. Second, I described the results of a new method to assess a reliable detection of unconscious stimuli in hemianopic patients probably triggered by evolutionary ancient subcortical areas. Finally, by using the Poffenberger paradigm, associated or not with ERP recordings, it was possible to have clues as to the importance of the intact hemisphere in mediating blindsight.