OK, first of all, I wish to thank the organizers who invited me to talk today. I enjoyed it very much. Let me introduce myself. I'm a clinician neuropsychologist and researcher in Montpellier, France. I work with Professor Hugues Dufault. So today I'm going to talk to you about new interoperative procedures used to map cognitive functions both at the cortical and subcortical level, especially in the right hemisphere. We have made a lot of progress the last two years. I shall begin by reminding that the goal of the OX surgery is to map and spare critical cortical epicenters for which the spatial topography differs from one patient to another, but also the critical cortical white matter connection in order to preserve range and critical function and consequently to preserve quality of life. The plasticity is generally strong, it is diffused by glioma patients, explaining why it is possible to perform radical resection in patients without inducing severe neuropsychological impairment. Here we have the example of 231 resection cavities of patients having resumed a normal social professional life. We have also the spectacular case of bifrontal lobectomy in a patient with only moderate cognitive impairment. However, plasticity is limited, most notably for a few structures in the brain, notably white matter connectivity, unimodal awareness, and a relatively small set of neural hubs, such as the posterior cingulate cortex. The question is how to map the critical structures. Sandor approached in OX surgery to map motor-osmotic sensory function to avoid problematic deficits after surgery. Here we see the distribution of cortical sites inducing articulatory disorders, split arrest, or facial deviation. Here is the distribution of subcortical sites involved in the motor-negative network. And here is the probability of observing an involuntary movement in the brain and the white matter fibers during electrostimulation. Of course, the mapping of spoken language processes is essential to avoid lasting and disabling post-operative aphasia disorders. Most of the time, a naming task is used. Depending on the structure simulated, a number of transitory impairments will be observed, such as anemia, semantic paraphagia, or phonological paraphagia. For example, we know that the probability of observing a semantic paraphagia in the white matter forming the left ventral connectivity is very high, naming the inferior frontal occipital fasciculus. It is common practice to use line bisection tasks to assess spatial cognition in order to avoid lasting post-operative spatial neglect. This task is useful to map the right posterior parietal cortex and the dorsal connectivity, especially the superior longitudinal fasciculus. We know that preserving the dorsal connectivity allows the patient to quickly recover after surgery, despite most of the time transitory impairment. So what about other functions? Do preserve spatial cognition, motor ability, or spoken language processing is enough to guarantee a good quality of life after surgery? My answer is definitively not. In this talk, I will show how we currently map a range of functions, notably mentalizing, nonverbal semantics, reading and executive and working memory, and conscious information processing. So first, why should we preserve mentalizing? Mentalizing refers to the set of processes involved in the understanding of social environment. Mentalizing allows people to make inferences about the mental state of others, such as intention, emotion, desires. Mentalizing is observed by a large frontoparietal network in the brain, and we know that mentalizing impairment can lead to aberrant social behavior, as in autism spectrum disorders or schizophrenia. For example, the case of this patient, who developed a paranoid delusion in connection with an impairment of mentalizing after a surgery involving the anterior insula and the parsopercularis of the right inferior frontal gyrus. Patients with glioma suffer from mentalizing impairment. We have previously shown that mentalizing impairment were correlated with the volume of resection of the right parsopercularis. We have also shown that mentalizing impairment were more frequent in patients with a resection of the parsopercularis and a resection of the most posterior part of the dorsolateral prefrontal cortex in the right hemisphere. In the operative theater, we use an adapted version of the Read the Mind in the Eyes task initially developed by Baron Cohen. This task consists in the presentation of photographs depicting the eyes region of human faces. Patients are asked to select among two propositions with which mental state best describes what the character is feeling or thinking. We advise the patient to make a self-judgment from one low confidence to six very high confidence. I show you here the time course of mapping. We always begin with a series of trials without stimulation to ensure that the patient performs the task appropriately. When we begin to stimulate, you receive a stimulation of the dorsal part of the parsofterior angularis and you see an increase of reaction time but not stimulation of the ventral part of the triangularis. In the same way, stimulating the lateral part of the right parsopercularis induces an error and an increase of reaction time but not other sites in the orbitalis, And finally, we observe in this patient that stimulating the dorsal triangularis and the lateral part of the opercularis induce frequently mentalizing impairment. We have used this protocol in five patients. In the illustration, the cavity resection is in red. The negative mapping is indicated by white diamonds and the positive mapping by the black arrows. We observe that the lateral part of the parsofterior angularis seems to be a critical cortical epicenter in the face-based mentalizing network but also the dorsal part of the parsofterior angularis. In two patients, we showed also that critical site in the white matter fiber underlying the white matter connectivity. Based on these results, we continued to map mentalizing processes in the white hemisphere in 28 other patients In agreement with our previous results, we showed that critical cortical sites in the parsofterior angularis and the parsofterior triangularis, we showed also numerous subcortical sites in the most posterior part of the dorsolateral prefrontal cortex but also at the level of the posterior part of the superior temporal gyrus. At the subcortical level, we found also a lot of responsive structure in the white matter fiber underlying both the dorsolateral prefrontal cortex but also in the white matter fibers underlying the middle temporal gyrus and the posterior temporal gyrus. The temporal stimulation were located along the spatial course of the white frontal occipital fasciculus. We can see here the anatomical dissection of the white HIFOF. As you can see, the white HIFOF interconnects the posterior occipital area and the posterior temporal area to the frontal cortex, especially at the level of the inferior frontal gyrus and the dorsolateral prefrontal cortex. If we replace the cortical sites on this dissection, approximately, we can see that temporal sites are indeed located on the spatial course of the HIFOF as well as the majority of the dorsolateral prefrontal cortex sites. However, we can see also that some sites are very posterior in the brain and does not match the spatial course of the HIFOF. So it is legitimate to ask the question, what is really disconnected during subcortical stimulation, notably in the posterior part of the dorsolateral prefrontal cortex? To provide an answer, we calculated for each subcortical stimulation point and disconnection map using the new software by Michel Chabot de Chauton. This search map indicates the probability of one miter fiber passing through the stimulation pipe or emanating from the stimulation pipe to the point to be disconnected. The yellow color in the illustration indicates that the probability is maximum. Then we averaged this disconnection map according to the location of the subcortical stimulation, and I mean the prefrontal of temporal. Using this method, we showed that subcortical stimulation are likely to disconnect the dorsal pathway, but also in lesser extent the white HIFOF. We showed also that the probability of the white HIFOF to be disconnected during temporal stimulation is very high. Therefore, it seems that both dorsal and ventral pathways are involved in face-based mentalizing. Beyond this fundamental implication, using a mentalizing test during awake surgery is useful on several levels. As the majority of lesions have an insular origin, the mentalizing test is useful to assess the functionality of the inferior frontal gyrus and may allow the neurosurgeon to decide where passing through the inferior frontal gyrus in the case of transopercular approach. The mentalizing test is also useful to determine where stopping the Y-section in the dorsolateral prefrontal cortex to large resection in this area can lead to a mentalizing impairment. Finally, the mentalizing task may help to find the deep end dorsal limits of temporal resection and the white HIFOF. The white ventral connectivity is involved in social cognition, but also in nonverbal semantic cognition. Previously, we have shown that the left HIFOF was involved in verbal and nonverbal semantic cognition using a simple picture naming task or visual semantic association task. The latter consisting in semantically matching two pictures. We used the same protocols to disentangle the role of the white hemisphere in verbal and nonverbal semantic cognition. Specifically, we assessed 13 patients presenting with diffuse low gray glioma in the frontal, insular, temporal, and parietal cortex. In almost all patients, we found eloquent sites for nonverbal semantics, but not for semantic cognition. Specifically, we found cortical sites in the posterior part of the prefrontal cortex, in the pars triangularis, and at the level of the superior temporal gyrus. In other words, in cortical area that mirror the semantic network in the left hemisphere. Most of the subcortical sites were located in the spatial cortex of the HIFOF. The two sites in the prefrontal cortex are probably also located at the level of the dorsolateral branch of the HIFOF, which is not represented on the used atlas of white matter. Therefore, semantic association task is useful to map and spare the white ventral connectivity. Let me talk about now from the interoperative making of reading with the frontal lobectomy in a patient with only moderate cognitive impairment. However, plasticity is limited, most notably for a few structures in the brain, notably white matter connectivity, unimodal awareness, and a relatively small set of neural hubs, such as the posterior cingulate cortex. The question is how to map the cortical critical structures. Sandor approached awake surgery to map motor somatosensory function to avoid problematic deficit after surgery. Here we see the distribution of cortical sites inducing articulatory disorders, split arrests, or facial deviation. Here the distribution of subcortical sites involved in the motor negative network. And here the probability of observing an involuntary movement in the brain in the white matter fibers during electro-stimulation. Of course, the mapping of spoken language processes is essential to avoid lasting and disabling post-operative aphasia disorders. Most of the time, a naming task is used. Depending on the structure simulated, a number of transitory impairments will be observed, such as anemia, semantic paraphagia, or phonological paraphagia. For example, we know that the probability of observing a semantic paraphagia in the white matter forming the left ventral connectivity is very high, naming the inferior frontal occipital fasciculus. It is common practice to use a line bisection test to assess spatial cognition in order to avoid lasting post-operative spatial neglect. This test is useful to map the right posterior parietal cortex and the dorsal connectivity, especially the superior longitudinal fasciculus. We know that preserving the dorsal connectivity allows the patient to quickly recover after surgery, despite most of the time transitory impairment. So what about other functions? Do preserve a spatial cognition, motor ability, or spoken language processing is enough to guarantee a good quality of life after surgery? My answer is definitively not. In this talk, I will show how we currently map a range of functions, notably mentalizing non-verbal semantics, reading and executive and working memory, and conscious information processing. So first, why should we preserve mentalizing? Mentalizing refers to a set of processes involved in the understanding of social environment. Mentalizing allows people to make inferences about the mental state of others, such as intention, emotion, desires. Mentalizing is observed by a large frontoparietal network in the brain, and we know that mentalizing impairment can lead to aberrant social behavior, as in autism spectrum disorders or schizophrenia. For example, in the case of this patient, we developed a paranoid delusion in connection with an impairment of mentalizing after a surgery involving the anterior insula and the pars opercularis of the right interferofrontal gyrus. Patients with glioma suffer from mentalizing impairment. We have previously shown that mentalizing impairments were correlated with the volume of resection of the right pars opercularis. We have also shown that mentalizing impairments were more frequent in patients with a resection of the pars opercularis and a resection of the most posterior part of the dorsolateral prefrontal cortex in the right hemisphere. In the operative theater, we use an adapted version of the read the mind in the eyes task initially developed by Baron Cohen. This task consists in the presentation of photographs depicting the eyes region of human faces. Patients are asked to select among two propositions with which mental state best describes what the character is feeling or thinking. We advise the patient to make a self-judgment from one low confidence to six very high confidence. I show you here the time course of the mapping. We always begin with a series of trials without stimulation to ensure that the patient performs the task appropriately. When we begin to stimulate, here we see the stimulation of the dorsal part of the pars opercularis induces an increase of reaction time. But not stimulation of the ventral part of the triangularis. In the same way, stimulating the lateral part of the right pars opercularis induces an error and an increase of reaction time. But not other sites in the orbitalis, et cetera, et cetera, et cetera. And finally, we observe in this patient that stimulating the dorsal triangularis and the lateral part of the opercularis induces frequently mentalizing impairment. We have used this protocol in five patients. In the illustration, the cavity rejection is in red, the negative mapping is indicated by white diamonds, and the positive mapping by the black arrows. We observe that the lateral part of the pars opercularis seems to be a critical cortical epicenter in the face-based mentalizing network, but also the dorsal part of the pars triangularis. In two patients, we showed also that a critical site in the white matter fiber underlying the white matter connectivity. Based on these results, we continued to map mentalizing processes in the white hemisphere in 28 other patients. In agreement with our previous results, we showed that critical cortical sites in the pars opercularis and the pars triangularis. We showed also numerous sites, subcortical sites, in the most posterior part of the dorsolateral prefrontal cortex, but also at the level of the posterior part of the superior temporal gyrus. At the subcortical level, we found also a lot of responsive structure in the white matter fiber underlying both the dorsolateral prefrontal cortex, but also in the white matter fibers underlying the middle temporal gyrus and the posterior temporal gyrus. The temporal stimulation were located along the spatial course of the white frontoccipital fasciculus. We can see here the anatomical dissection of the white eye faux. As you can see, the white eye faux interconnects the posterior occipital area and the posterior temporal area to the frontal cortex, especially at the level of the inferior frontal gyrus and the dorsolateral prefrontal cortex. If we replace the cortical sites on this dissection, approximately, you can see that temporal sites are indeed located on the spatial course of the eye faux, as well as a majority of the dorsolateral prefrontal cortex sites. However, we can see also that some sites are very posterior in the brain and does not match the spatial course of the eye faux. So it is legitimate to ask the question, what is really disconnected during subcortical stimulation, notably in the posterior part of the dorsolateral prefrontal cortex? To provide an answer, we calculated for each subcortical stimulation point and disconnection map using the new software by Michel Thiebaud de Chauton. This such map indicates the probability of one miter fiber passing through the stimulation pipe or emanating from the stimulation pipe to the point to be disconnected. The yellow color in the illustration indicates that the probability is maximum. Then we average this disconnection map according to the location of the subcortical stimulation, and I mean the prefrontal of temporal. Using this method, we showed that subcortical stimulation are likely to disconnect the dorsal pathway, but also in lesser extent the white eye faux. We showed also that the probability of the white eye faux to be disconnected during temporal stimulation is very high, therefore it seems that both dorsal and ventral pathway are involved in face-based mentalizing. So beyond this fundamental implication, using a mentalizing test during eye surgery is useful on several levels. As the majority of lesions has insular origin, the mentalizing test is useful to assess the functionality of the inferior frontal gyrus and may allow the neurosurgeon to decide where passing through the inferior frontal gyrus in the case of transopercular approach. The mentalizing test is also useful to determine where stopping the resection in the dorsolateral prefrontal cortex. Too large resection in this area can lead to a mentalizing impairment, and finally the mentalizing test may help to find the deep endorsal limits of temporal resection, and I mean the white eye faux. The white ventral connectivity is involved in social cognition, but also in nonverbal semantic cognition. Previously we have shown that the left eye faux was involved in verbal and nonverbal semantic cognition using a simple picture naming task or visual semantic association task. The latter consisting in semantically matching two pictures. We used the same protocols to disentangle the role of the white hemisphere in verbal and nonverbal semantic cognition. Specifically, we assessed 13 patients presenting with a diffuse low-grade glioma in the frontal insular temporal and parietal cortex. In almost all patients, we found eloquent sites for nonverbal semantic, but not for semantic cognition. Specifically, we found cortical sites in the posterior part of the prefrontal cortex, in the pars triangularis, and at the level of the superior temporal gyrus. In other words, in cortical areas that mirror the semantic network in the left hemisphere, most of the subcortical sites were located in the spatial cortex of the eye faux. The two sites in the prefrontal cortex are probably also located at the level of the dorsolateral branch of the eye faux, which is not represented on the used atlas of white matter. Therefore, semantic association tasks is useful to map and spare the white ventral connectivity. Let me talk about now from the interoperative making of reading. Reading is a crucial function in daily life and we have to preserve it in patients. Reading is made possible through very distributed processes in the brain, very schematically, the visual information is entered in the occipital pole and quickly moves towards a very specific area called the visual word form area, which is specialized in the recognition and letter and the white and words. Then the neural information processed by the visual word form area is distributed in at least two networks. One network is involved in reading and the second network is involved in the pronunciation and articulation of the word. So, how to map this core network for reading during AWACS surgery? In the operating theater, we use a simple reading task. Specifically, we ask a patient to read a word every four seconds. There are different kinds of words to be read, regular words, irregular words, and pseudowords. The key idea is to map the different kinds of processes involved in reading. We use this task in seven patients with diffuse logarithm located near or close the visual word form area and we showed that simulating the visual word form area induces different kinds of alexia depending on the site of stimulation. Specifically, we showed that simulating the posterior aspect of the visual word form area induces complex alexia, naming a disturbance of reading regular, irregular, and pseudowords. So, we showed also that cutting the posterior part of the inferior longitudinal fasciculus interconnected the occipital pole in the visual word form area induced also complete alexia. Simulating the dorsal aspect of the visual word form area induced an alexia for irregular and pseudowords, but also when cutting the posterior part of the arcuate fasciculus. In the same way, simulating the anterodorsal part of the visual word form area induced an alexia for regular words, and finally, cutting or stimulating the anterior part of the inferior longitudinal fasciculus has no effect on reading. In our center, we use also other tasks. To assess executive processes, we ask the patient to make a dual task consisting as a regular movement on both the upper arm and the upper hand in concert with a naming task or a semantic matching association task. This allows to assess at the same time language, motor movement, executive processes, motor movement, executive function, and working memory. In the case of a patient with a high educational level, we rather use a handbag task consisting in naming the image seen one trials or two trials before. For example, the patient will say twain instead of dwam, etc. To conclude, there are two particular points I would like to underline. First, it is my opinion that spoken language and sensorimotor function is not sufficient to preserve quality of life. Other tasks should be used to map a number of critical functions. Second, the right non-dominant hemisphere is eloquent for other functions of spatial cognition. Regarding this later point, the case of this white-handed patient with posterior prefrontal rejection is exemplary by demonstrating the functional necessity to a white patient with a white hemispheral tumor. We have indeed observed in this patient numerous cortical and subcortical sites, anartrias, speech airways and movement eruptions, dystonic movement, involuntary movement by stimulating the premotor cortex and motor cortex, phonological disturbances by stimulating the parts of percolaris, semantic impairment and mentalizing impairment by simulating the most posterior part of the posterior prefrontal dorsolateral prefrontal cortex at the subcortical level, articulatory disorders, involuntary left saccades, singing, and mentalizing impairment. A frequently asked question is to know whether we should use other tasks during the intraoperative med dealing. This is a valid question, but we can only answer it by systematically performing comprehensive neuropsychological examination before and three months after surgery and by comparing this examination to identify which function do recover and which function do not recover. We have also to determine the extent to which impaired function actually impact quality of life. Thank you for listening.

brain surgery

cognitive functions

plasticity

mentalizing

awake surgery

language functions

motor skills