The perspectives of mapping and monitoring of the sense of self in neurosurgical patients

Abstract

Surgical treatment of tumors, epileptic foci or of vascular origin, requires a detailed individual pre-surgical workup and intra-operative surveillance of brain functions to minimize the risk of post-surgical neurological deficits and decline of quality of life. Most attention is attributed to language, motor functions, and perception. However, higher cognitive functions such as social cognition, personality, and the sense of self may be affected by brain surgery. To date, the precise localization and the network patterns of brain regions involved in such functions are not yet fully understood, making the assessment of risks of related post-surgical deficits difficult. It is in the interest of neurosurgeons to understand with which neural systems related to selfhood and personality they are interfering during surgery. Recent neuroscience research using virtual reality and clinical observations suggest that the insular cortex, medial prefrontal cortex, and temporo-parietal junction are important components of a neural system dedicated to self-consciousness based on multisensory bodily processing, including exteroceptive and interoceptive cues (bodily self-consciousness (BSC)). Here, we argue that combined extra- and intra-operative approaches using targeted cognitive testing, functional imaging and EEG, virtual reality, combined with multisensory stimulations, may contribute to the assessment of the BSC and related cognitive aspects. Although the usefulness of particular biomarkers, such as cardiac and respiratory signals linked to virtual reality, and of heartbeat evoked potentials as a surrogate marker for intactness of multisensory integration for intra-operative monitoring has to be proved, systemic and automatized testing of BSC in neurosurgical patients will improve future surgical outcome.

Keywords: Bodily self; Brain surgery; Consciousness; Neuropsychology; Self; Social.

Fig. 1

Self-other voice discrimination task: Each participant’s voice was recorded prior to the experiment while vocalizing phoneme /a/ for 2 s. Participant’s voice (the orange bar in a) was than morphed with a voice of a gender-matched unfamiliar person (the blue bar in a) in order to generate a self-other voice identity continuum. From that continuum, six voice morphs (% self-voice 15, 30, 45, 55, 70, 85) were presented randomly to participants while recording the electrophysiological activity with a high-density EEG cap (lilac spheres and connections in b). The cap is formed by 256 electrodes organized as an extension of the standard clinical 10–20 setup (whose electrodes names are indicated in black in b). After hearing a voice morph, participants were asked to indicate whether the voice they heard sounded more like their own or someone else’s by pressing the corresponding mouse button (c). Voice morphs were presented either through laptop loudspeakers (not illustrated) or a bone-conduction headset (illustrated in black in b)

Fig. 2

Localization of the self-voice: The analysis of the EEG during the voice task discrimination allows to define the network that significantly and specifically activates when participants hear their own voice. The results are visualized on a MNI template, and the results of activation have been obtained by projecting in the “brain space” (inverse space) the EEG signal acquired on the scalp, with the academic free Cartool software (https://sites.google.com/site/cartoolcommunity/). The brain network of the self-voice includes the insulae and putamen and the maximum of activation is lateralized on the right hemisphere (red crosses on the brain images in the red boxes). Moreover, the network includes the middle cingulum and part of the right inferior temporal pole [from Orepic et al., in submission]

Fig. 3

Heartbeat evoked potentials (HEPs). Mechanoreceptors on the heart wall discharge at a specific phase of EKG. Visceral information is then relayed up through cranial nerves and spinal relays to cortical and deep structures, among which the amygdala, the region of the ventral anterior cingulate cortex-ventral anterior prefrontal cortex (vACC-vmPFC), the insula, and the somatosensory cortex (SS cortex) play an important role in the integration of these signals. HEP can then be recorded in this region just after the R-wave on the EKG (electrocardiogram). Park et al. Cereb Cortex [91], Armour and Ardell [5]

Fig. 4

Self-related HEPs. a BSC modulations which were experimentally induced using synchronous or asynchronous visuo-tactile stimuli were associated with the HEP amplitude at the fronto-central region. A follow-up intracranial EEG study confirmed that self-related HEP can be measured at the insular cortex (b). HEP heartbeat evoked potential

Fig. 5

HEP during anesthesia. a During operation, intracranial EEG signals were recorded from contacts on the posterior insular cortex. b Spectral power of HEP from the posterior insular cortex. Increased power was observed around 200 ms after the ECG R-peak onset, from 5- to 10-Hz range. HEP heartbeat evoked potential