In WP1, I will conduct longitudinal studies in foetuses, newborns and infantsto describe the emergence of implicit biomarkers of three BSR components.

Spatial Location (SL): This first component, intended asthe representation of the bodily-self in space, can be accessed either by proprioceptive (SLp) or by visual (SLv) channels. Neuropsychological evidence informs us that a proprioceptive representation of the bodily-self in space has a pivotal role for the correct functioning of BSR since, due to position sense loss, it is always prevented in BSR disorders, such as pathological embodiment, and somatoparaphrenia. This component will be investigated to explore the emergence of a primitive coding of the bodyself boundaries in prenatal life and its specialization along the postnatal period, by exploiting the spatial modulation of multisensory integration (MSI), which is considered a hallmark of a functioning representation of the body position in space. Sensitivity to multisensory correspondences underpinning body representations is available early in postnatal life and, as I contributed to describe, EEG markers of a spatially organized MSI can be measured soon after birth. However, no previous studies investigated MSI in prenatal life. Here, I propose innovative methodologies to investigate whether MSI is already present in foetuses and spatially modulated by the proximity to the foetal body. Furthermore, I will exploit a MSI paradigm to investigate when and how, during postnatal life, the multisensory encoding of the body position in space is circumscribed around a specific effector (i.e., the hand).

When proprioception is lost, patients could vicariously exploit visual input to localize their own body in space (SLv), as shown by the presence of a spatially organized MSI when the affected hand is visible12. The ability to localize the arm position by sight has been described in non-human and human primates, with key brain regions that are modulated by visuo-proprioceptive congruency. However, when this ability develops along the postnatal life is still unknown. I will investigate the development of the visual localization of the bodily-self in space by describing when and how visuo-proprioceptive congruency is able to modulate MSI, as I previously demonstrated in adults.

Prototypical Configuration (PC): The combination of visuo-proprioceptive signals forms the basis of PC, intended as the representation of the bodily-self prototypical configuration in space. Due to a spared visual channel, E+ patients strongly rely on this prototypical body representation. Indeed, they misidentify the alien hand as their own only when it matches the constraints of this prototypical representation (e.g., a human hand aligned to the shoulder and perceived in first-person perspective). In developmental studies, the role of visual experience in the development of this component has been investigated by exploiting the crossed-limbs deficit in tactile localization, as a marker of the representation of the body canonical layout in external space. In congenitally blind adults, the absence of the crossed-limbs deficit suggests that the emergence of this prototypical body representation is prevented by visual deprivation. By contrast, in sighted infants, orienting responses to tactile stimuli suggests that the crossed-limbs deficit arises early in life. Here, the development of this component along the postnatal life will be investigated by describing when and how prototypical constraints are able to modulate MSI.

Visual Identification (VI): While the role of the two previous components has been investigated also in classical body illusions in healthy subjects, less attention has been paid on VI, intended as the representation of the identity-related visual features of the bodily-self. A damage of this component represents the core deficit of pathological embodiment, being selectively impaired in E+ patients who completely ignore body-identity visual details (i.e., skin colour, shape, age or dimension) and fail in self/other visual discrimination. In the developmental context, previous studies focused on the visual recognition of the face, given its pivotal role for the human identity. Here, I will focus on the visual recognition of the hand, as a neglected yet fundamental line of research. Representing a preferential mean of interaction with the environment and with our peers, the hand has a crucial role in both phylogenetic and ontogenetic human development.The emergence of the self-hand visual identificationwill be investigated postnatally, when babies start staring at the own hand as a functional step to develop visually-guided action, by comparing responses to self and other hand images. 

From the results of WP1, I expect to describe a clear picture of the crucial phases of the BSR development, starting from a primitive coding of the bodily-self in space, which likely emerges in the maternal womb, and proceeding to further specializations along the post-natal life, from a spatial coding around body effectors to the maturation of an abstract knowledge of the prototypical configuration of the bodily-self and of its identity-related features. 

In WP 2, I will exploit neuroimaging techniques to investigate the causal role of key brain regions and of their functional connectivity in guiding the emergence, along prenatal and postnatal life, of the BSR components at specific timepoints identified by longitudinal studies in WP1. 

Spatial Location (SL): Converging evidence from neurophysiological studies in monkeys and neuroimaging studies in humans show that the multisensory representation of the bodily-self in space is mediated by both frontal (ventral premotor cortex, PMv) and posterior parietal (PPC) regions. Such areas house multisensory neurons that preferentially respond to stimuli occurring close to (as compared to far from) the body, by increasing their activity and their functional connectivity. In E+ patients, a spared activity of these brain region can support multisensory functions when information about the body position in space is accessed via the visual (instead of proprioceptive) channel12. Importantly, in developmental context, a previous study exploiting resting-state fMRI in newborns suggests that the functional architecture needed for MSI is already present at birth. Here, I will investigate whether the amount of activation over these multisensory areas, as well as their increased connectivity, may predict the extent to which foetuses and infants are sensitive to body proximity and (for infants only) to visuo-proprioceptive congruency, as revealed by the effectiveness of these manipulations in modulating MSI. 

Prototypical Configuration (PC): Converging evidence from BSR manipulation in healthy subjects show that the prototypical knowledge of the canonical body configuration is mediated by the activity of the temporo-parietal junction (TPJ). This brain region is known to play a crucial role in promoting a “test for fit”, by selecting what can be accepted as a part of the own body based on specific constraints. Importantly, in E+ patients who strongly rely on this BSR component and incorporate the alien limb only when it fits the prototypical constraints, TPJ is usually spared in the lesion mapping. Here, I will investigate whether the amount of activation over TPJ, as well as its increased connectivity with multisensory regions, may predict the extent to which infants are sensitive to prototypical constraints, as revealed by the effectiveness of prototypical constraintsin modulating MSI. 

Visual Identification (VI): In a recent tractography study from my Lab, I showed that, in E+ patients, the third branch of the superior longitudinal fasciculus, connecting the PMv to the inferior parietal cortex, and the posterior arcuate fasciculus, conveying information coming from occipito-temporal areas to parietal regions, are the main fibre tracts involved. Thus, a pathological bodily-self visual recognition, as the core deficit of E+ patients, emerges from the disconnection between visual (computed in occipito-temporal areas) and sensorimotor (stored in premotor and parietal areas) body representations. Accordingly, also different lines of research investigating BSR manipulations in healthy subjects proposed that the normal functioning of such visual-sensorimotor network is critical to recognize one’s own body. Here, I will investigate whether the amount of activation over visual and sensorimotor areas, as well as their reciprocal connectivity, may predict the extent to which infants are able to discriminate self and others hand images. 

From the results of WP2, I expect to describe the network dynamics underling the BSR development, starting with the maturation of multisensory areas and their functional connectivity along the prenatal and the postnatal life to the emergence of specific brain networks supporting more abstract representations of the bodily-self. 

In WP3, to challenge the causative role of the motor context in the emergence of BSR, I will capitalize on congenital motor deprivation, i.e., cerebral palsy (CP). The deprivation of the motor context since the prenatal life should prevent the association between visual, tactile and proprioceptive inputs, leading to an impairment of the BSR development. Therefore, I will recruit a sample of CP (adult) individuals who suffered a prenatal brain insult, resulting in severe contralateral upper-limb hemiplegia. While previous studies extensively investigated the role of visual deprivation in the BSR development, less attention has been paid to motor deprivation. Besides, previous studies in CP focused mostly on the disabling motor deficit. However, recent evidence suggests that these patients may show an altered body representation as well, demonstrating a growing interest on this topic, also in clinical realms. In WP3, by exploiting the very same methodologies of WP1 and WP2, I will test whether the BSR components and their underlying neural mechanism are affected in congenital motor deprivation (i.e., CP individuals with upper-limb hemiplegia), as compared to acquired motor deprivation (i.e., post stroke patients with chronic upper-limb hemiplegia). 

Spatial Location (SL): The multisensory representation of the body in space is strongly related to the motor context, since this representation integrates external stimuli (visual, auditory), which potentially interact with the body (when occurring close to it), with somatosensory cues (tactile, proprioceptive), thus driving potential motor responses. In the social context, such as during affective touch, the repeated experience of seeing and feeling one’s body being touched will link visual and tactile representations. Therefore, CP individuals might develop the association between visual/acoustic and tactile stimuli that, in turn, is crucial to show a MSI effect. However, the motor context, wherein infants see their own hand touching an object during reaching or grasping, is crucial to associate also proprioceptive inputs. Such association is crucial to show a spatially organized MSI. Indeed, multimodal neurons encode multisensory bodily stimuli in body-centred reference frames by anchoring visual, auditory and tactile receptive fields to a given body part by means of proprioceptive inputs. Patients with acquired motor deprivation can access the spatial representation of their affected body side, at least when they vicariously exploit vision instead of (often impaired) proprioception. Accordingly, previous studies in short and long immobilization showed that motor deprivation affects the bodily-self spatial representation when visual information is precluded. Contrarily, I predict that CP individuals fail in accessing SL, irrespective of which channel they use (either proprioceptive or visual), since multisensory stimuli have not been associated to proprioceptive inputs through developmental motor experiences.

Prototypical Configuration (PC): In the BSR literature, there is a strong agreement about the sensorimotor nature of this component. Accordingly, the intersensory contingency between visual information from one’s own body posture and tactile and proprioceptive inputs during movements may lead infants to build prototypical representations of the body, which in turn define what can/cannot be accepted as belonging to their own body . While patients with acquired motor deprivation can rely on the already structured prototypical configuration, in CP individuals, the congenital motor deprivation should prevent the maturation of prototypical constraints through development, as may be proved by the ineffectiveness of prototypical constraints in modulating MSI in the affected side. 

Visual Identification (VI): While we usually distinguish other people’s bodies by vision only, it has been proposed that for bodily-self recognition we also rely on sensorimotor representations. The additional recruitment of a sensorimotor network is thought to mediate the bodily-self visual recognition and its EEG correlates that I described previously. In the first 6 months, since infants engage in “pre-reaching” movements a lot, they intensively gaze at their hands, thus associating the sensorimotor feedback with the hands’ visual features. While some patients with acquired motor deprivation can rely on the already established visual-sensorimotor association, in others (i.e. E+ patients), the extent of the damage to the visual-sensorimotor network predicts the severity of the bodily-self visual recognition impairment33. In CP individuals, the congenital motor deprivation should preclude the association between visual and sensorimotor representations of the self-hand in the affected side, thus preventing the full maturation of the visual-sensorimotor network that sustains visual identification. From the results of WP3, I expect to provide the proof that early motor experience represents the crucial context for the association among multisensory signals, which leadsto the gradual emergence of BSR.