Thank you very much, and many thanks to the organizers for having invited me to this really exciting meeting. So I'm not a neurosurgeon. I'm a neurologist particularly interested in the consequences of brain damage on visual function, especially high-level vision. So I hope to, while I have no hope to cover the entire visual function and spatial cognition in 20 minutes, perhaps we can discuss together about a few instances relevant to neurosurgery. So first basic distinction in vision in the brain is that between low-level vision, so basically vision between the retina and V1, the primary visual cortex, and high-level vision, everything which happens after V1. So low-level vision is already a clinically relevant issue, but Hugues Dufour has already explained how it is possible with a wake surgery to prevent, for example, hemianopia or to reduce the severely disabling hemianopia to, for example, a quadrantic visual deficit, which is much less disabling. Let us discuss high-level vision. Here the traditional distinction in cognitive neuroscience is that between two cortical visual pathways, the well-known distinction between the Watt pathways or ventral occipital temporal pathways, mainly devoted to object recognition, to visual recognition in general, and a more dorsal occipital parietal prefrontal pathways, which is important for more specially related activities, such as localizing objects in space or programming accurate hand movements towards objects in space. So let's start from considering the ventral occipital temporal stream. Here you can see the odology of this stream, which has been already covered today. The main pathways are the ILF, the inferior longitudinal fasciculus here in green, and the IFOF, the inferior frontal occipital fasciculus. For some other functions also, as we have heard, the ancinate fasciculus could be important. What fMRI studies have shown concerning this ventral occipital temporal pathway is that there is a set of regions which are relatively specialized for a specific object categories. Typically, there are regions specialized for face recognition, regions specialized for place recognition, as you can see here. There is the so-called the visual work form area, which is important for reading. There are other regions not shown here, important for color processing and for movement processing. Don't get me wrong, I'm not saying that all the relevant processing concerning these categories is happening in these specialized regions. These regions are to be conceived as important nodes in networks, for the identification of this particular object features. From the clinical point of view, the important thing is that most of these regions are bilateral. So, unilateral damage to, for example, the LOC, important for object shape, is in principle should not give rise to object agnosia. You need bilateral lesions to give rise to object agnosia. One relevant exception to this pattern is the visual work form area, which is important for reading, for identifying for identifying orthographic material. Here, you should be able to see. No, strange, it doesn't work. Wow, okay. This was an... Okay, here you can see it. These are MEG data, magnetoencephalography data, showing the progression of perception of words from the occipital pole to the temporal pole and then to the language areas during reading. So, we have already heard from you before about reading. Historically, the first case of pure Alexia, of selective deficit, acquired deficit of reading, was described by Jules de Chérin, who described this stroke patient who, from one day to another, lost all capacity of reading, even single letters. Lost all capacity of reading, even single letters and single numbers. It was an extremely severe form of Alexia, and it traced down de Chérin this deficit to the fact that the region important for what he called the visual images of word, which he incorrectly localized in the angular gyrus, was disconnected from the visual images of word, which he incorrectly localized in the angular gyrus, was disconnected from the visual images of word, which he incorrectly localized in the angular gyrus, from the primary visual cortex in the left hemisphere, but also from the primary visual cortex in the right hemisphere because of the disconnection of the splenium of the corpus callosum. This is a more recently studied patient, neurosurgical patient. He was a patient studied by the group of Laurent Cohen in the Salpeteriere Hospital in Paris, in my lab actually. This patient had a drug-resistant epilepsy, and so he was, he received a surgical treatment for this epilepsy, not awake surgery, and surgery interrupted the fibers of the inferior longitudinal fasciculus, which you can see here, and as a result, the visual word form area, which was identified by using fMRI, was disconnected by V1, and the patient became alexic after the operation. Happily, his alexia was much less severe than that of the Jorin's patient, and it actually was some kind of letter-by-letter reading, rather than pure, complete alexia, and the patient actually recovered to a good extent after a few months, probably because the fibers of the splenium of the corpus callosum were spared by the intervention, and so the visual word form area could receive probably visual input from the right hemisphere. So let's discuss now the occipital parietal prefrontal stream, the so-called dorsal visual stream, here. Functional neuroimaging has identified at least two networks, important for the orienting of attention, of visual attention, in space. So according to this fMRI data, there is a dorsal attention network consisting of the inferior parietal sulcus and the superior parietal lobule, and of the frontal eye field. This bilateral network is important for orienting attention in space. So if you orient your attention left or right, you are using this bilateral dorsal attention network. But the Corbetta group discovered also a more ventral network consisting of the TPJ, the temporal parietal junction, so the inferior parietal lobule, and the caudal part of the superior temporal gyrus, and of the inferior and middle frontal gyrus. This network, which interestingly is especially represented in the right hemisphere, so in the hemisphere non-dominant for language, has a its function is more difficult to define, but for example, it activates when you have to interrupt an inappropriate orienting, and to reorient your attention to the appropriate object. For example, if you are crossing the street and then a car appears coming towards you at high speed, you have to, of course, interrupt your previous orienting of attention, reorient towards the car in order to avoid it. In this case, this ventral attention network, once again, which is especially evident in the right hemisphere, activates in order to interrupt your inappropriate previous orienting, and reorient your attention towards the relevant but unexpected incoming object. This is the odology of these attention networks, as established by Michel Thiebaud de Schotten in the Catani lab. So Michel discovered that the dorsal attention network is linked by the so-called SLF-1, by the first branch of the superior longitudinal fasciculus, here in red. Then the ventral attention network is connected by the SLF-3, here in green. And then there is an intermediate part, branch, of the superior longitudinal fasciculus, the so-called SLF-2, which actually allows the two networks to talk to each other, because it connects the TPJ, so the caudal node of the ventral attention networks, to the frontal eye field, so the the dorsal, the anterior node, of the dorsal attention network. One important thing, one important result found by Michel Thiebaud in this study, is that there is some hemispheric asymmetry, anatomical hemispheric asymmetry. There is actually a gradient of hemispheric asymmetry, such that the the SLF-1 is relatively, is actually symmetrical. The SLF-2 presents some degrees of asymmetry in favor of the right hemisphere, but not in every subject, in some subjects. There is a larger SLF-2 in the right hemisphere than in the left, but not in all the subjects. And in in all the subjects, the SLF-3 is definitely larger in the right hemisphere than in the left hemisphere. So, with this anatomical asymmetry, it goes with the functional asymmetry shown by the fMRI studies, in showing that, indeed, there is something special in this ventral attention networks, which network, which happens only in the right hemisphere. So, as a consequence, I think we shouldn't anymore call the left hemisphere the dominant hemisphere, and the right hemisphere the non-dominant hemisphere. Let's try and qualify this by saying, at least, left hemisphere dominant for language, right hemisphere non-dominant for language, because there are other functions, less well-defined, perhaps, than language, for which the right hemisphere seems to be specialized. And the clinical issue most relevant with respect to these attentional networks is that, when there is dysfunction of these networks, you may, in the right hemisphere, you may observe, you usually observe, signs of left visual neglect. So, these patients behave as if the left part of the world did not exist anymore. Because, look at this patient, who's performing here a line cancellation task. So, you should find every target in the sheet, but look at how he concentrates his attention only on the right part of the sheet. So, this patient, these patients are, as you can imagine, severely disabled. They cannot have an independent life, because everything which happens on their left, they simply will not notice. Not because they don't see it, but just because they do not pay attention to everything which happens on their left. They eat, for example, only the food on the right part of the dish. They only answer to people who ask them questions from the right side, but not from the left side. They may even neglect to shave the left part of their face. So, they can neglect everything, including, in some cases, their own body, which is on the left part. So, it is a, as I said, a severely disabling deficit. If you don't know it, it's difficult to, or impossible, to make a diagnosis of neglect, because it is not like language problems, where you simply, the patient doesn't talk, or it doesn't understand what you say. It is mainly composed of negative signs, patients who do not pay attention to left-sided items. However, the good news is that a few paper and pencil tests, very easy to perform even at bedside, such as this one, can, can easily demonstrate signs of neglect. And neglect is also clinically important, because we know that those patients recover, I'm talking of stroke patients now, recover less well from motor deficits, for example. They do not pay attention to the left side of their body. They are often anosognosic. They do not, do not have a, are not aware of their condition, so they are not motivated to follow rehabilitation. And another test, which is very simple, very easy to do with this patient, is line bisection. You simply give the patient a horizontal line drawn on a sheet of paper, and ask the patient to mark the midline, the center of the line. Typically, these patients will mark the center on the right of the geometrical center of the line. And thanks to this test, we were able, with Michel Thiebaud and Hugues Dufault, to perform intraoperative remapping of visospatial functions in, in two patients. And because line bisection can be easily be performed in the operatory theater. So these are the results. When, when Hugues stimulated several regions traditionally considered important for neglect, so basically the, the caudal nodes of the ventral attention network, the, the, the caudal part of the superior, superior temporal gyrus, and the, the inferior parietal lobule, the patient did deviate, produce small but significant deviations towards the right during line bisection. However, when Hugues stimulated the bottom of the operatory cavity, so in contact with the white matter systems at the bottom of the inferior frontal lobule, inferior parietal lobule, then the deviations toward the right were massive. Such as, sorry, how are we doing? Okay. Such, such, such the deviation produced by patients with stroke and very severe neglect. So we concluded that, on the basis of this and many other data, that neglect is not caused by parietal damage, as it is often written in neurology textbooks. Neglect is, visual neglect is a network deficit, network-based deficit. So, resulting from the dysfunction of the attentional networks, which I presented you before. And when Michel Thiebaud performed DTI tractography on the brain of this particular patient, we found that the, the affected white matter tracts corresponded to the SLF2, that is to the tract, which makes the two attentional networks talk to each other. Importantly, this patient, well, the day after the intervention, he had massive neglect and a whole series of neurovisual symptoms, such as optic ataxia. But five days after, he had completely recovered. It didn't show any sign of neglect. So we proposed the use of line bisection as an intraoperatory mapping test for, for visual neglect, to prevent the occurrence of visual neglect. And I was very happy to see more recently this paper by the Milan group, including Lorenzo Bello, who is here, who were able to substantially confirm this evidence, showing right, right-sided shift, rightward shifts in line bisection during transitory inactivation of the SLF2. What happens when the SLF2 is cut instead? Unfortunately, we know what happens, because this team of Japanese neurosurgeons did not awaken their patient and they cut the SLF2, and as a consequence, both of these two patients presented, one, the occurrence of neglect, and the other, the worsening of a pre-existing neglect. So we were able to extend these findings also to to a population of patients with stroke, and actually we found that these are 58 patients with stroke. We found that damage to the SLF2 is the best predictor, actually, of neglect in this group of patients, followed by less statistically robust result, but also probably a reliable result of SLF3 disconnection, damage to the SLF3, and indeed this is a patient with a totally subcortical stroke, which impaired the SLF3, and he did have severe neglect in the acute phase without any cortical lesion. This is my last slide. This is another group of patients where we found, so again, this connection of the SLF3 in correlating with neglect, but also in a subgroup of patients, there was a damage to the right-sided IFOV, which could also be important for visospatial cognition. These are data to be confirmed. I hope that with you, we can test the functions of the IFOV in the operatory theater. So, take-home messages. First, I would say the left hemisphere is not the minor hemisphere. I hope I've convinced you that it's important. It is dominant for important attentional functions. Second, intraoperatory mapping of visual visual abilities should be performed, and it is relatively easy to perform, although much work remains to be done. Finally, well, but I think this is for this audience is trivial, even if awake surgery is not possible, we should try and respect and spare the long-range longitudinal white matter fibers, and also, I did not have time to to show this data, but we have evidence we published last year in InBrain that in stroke patients, the condition of the splenium of the corpus callosum could be important for recovery. So, those patients who had the completely spared splenium of the corpus callosum, had more chances to recover from neglect than those patients with the damage to the splenium of the corpus callosum. So, if we can spare also the splenium of the corpus callosum to preserve visual function, perhaps patients would have a better, will have a better outcome. So, I would like to thank my co-workers and especially Michel Thiebaud, my former, who was a former PhD student of mine, now is a friend and co-worker, and thank you for your attention.

neurologist

brain damage

visual function

high-level vision

low-level vision

object recognition