Hemispherotomy: cortical islands of deep sleep in awake humans

doi:10.21203/rs.3.rs-4314125/v1

Download PDF

Article

Hemispherotomy: cortical islands of deep sleep in awake humans

https://doi.org/10.21203/rs.3.rs-4314125/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Hemispherotomy is a surgical procedure that disconnects a large portion of the cerebral cortex from cortical and subcortical inputs in patients with severe refractory epilepsy. Whether the disconnected cortex - inaccessible to behavioral assessment - supports consciousness remains unknown. Functional MRI studies have indicated preserved resting-state networks within the disconnected hemisphere, raising the possibility that it may represent an ‘island of awareness’. However, these networks can also persist in unconscious states, such as anesthesia and deep sleep.

Here we assess the capacity of the disconnected cortex to support consciousness by exploring its electrophysiological state, before and after hemispherotomy, in ten awake pediatric patients.

After surgery, the disconnected cortex–but not the contralateral cortex–entered a state dominated by slow oscillations (<2 Hz) resembling those observed during deep sleep; further, the spectral exponent, a previously validated marker of consciousness indexing the 1/f-like decay of the power spectral density, assumed values typically found in unconscious brain-injured and anesthetized adults. When compared to a reference pediatric sample, spectral exponent values were compatible with wakefulness in the contralateral cortex but attained levels typical of deep sleep over the disconnected cortex, suggesting that the disconnected cortex is not an island of awareness.

Biological sciences/Neuroscience/Neural circuits

Health sciences/Neurology/Neurological disorders/Brain injuries

Biological sciences/Physiology/Neurophysiology

Biological sciences/Neuroscience/Neural circuits

Health sciences/Neurology/Neurological disorders/Brain injuries

Biological sciences/Physiology/Neurophysiology

EEG

1/f noise

slow wave activity

disconnective surgery

consciousness.

Hemispherotomy is a surgical procedure used to treat severe cases of refractory epilepsy in children^1,2. The procedure aims at maximal disconnection of the pathological cortex from the rest of the brain by severing its white matter connections with brainstem, basal ganglia, thalamus and contralateral hemisphere. Typically, the extent of the cortical disconnection in hemispherotomy ranges from one lobe to an entire hemisphere. The disconnected/isolated cortex, completely devoid of communication with the external environment (except for cortical afferents from the olfactory bulb, in some cases) is left inside the cranial cavity with preserved vascular supply³. This neurosurgical manipulation is effective in blocking the spread of epileptic seizures from the pathological cortex to the rest of the brain, allowing patients to lead satisfying lives ^2,4,5. It is clear that the intact contralateral cortex (which remains connected to the body and the external world) continues to support consciousness after surgery, as patients regain (at least) partial sensory-motor functions, cognitive abilities and the capacity to report their internal states ^2,6. However, the disconnected cortex remains inaccessible to behavioral evaluation, due to the sensory and motor disconnection; thus, assessing whether it retains some form of awareness is challenging and must solely rely on brain-based measurements ^7,8.

The available evidence is drawn from two fMRI studies. The first study, involving two pediatric participants, revealed preserved lateralized connectivity in resting state networks within the disconnected hemisphere, despite decreased regional cerebral blood flow⁹. These preliminary results have recently been confirmed in a larger cohort¹⁰. This study found preserved functional connectivity in all resting-state networks, including the default mode network (DMN), in the disconnected hemisphere, albeit with reduced between-network segregation. The functional integrity of resting-state networks is compatible with the possibility that the disconnected cortex may constitute an ‘island of awareness’¹⁰. On the other hand, the partial preservation of resting state networks, including the DMN, does not necessarily indicate the presence of consciousness. In fact, residual functional connectivity within these networks can be preserved in unconscious conditions, such as coma, general anesthesia, and deep Non-Rapid Eye Movement (NREM) sleep^11,12,13,14.

For this reason, it is important to investigate further the state of the disconnected cortex by a complementary exploration of cortical activity. A large body of literature in healthy and pathological brains suggests that reliable indexes of loss and recovery of consciousness can be obtained from the analysis of the EEG^15,16,17,18. Specifically, a long-standing observation in cases of reduced or suppressed consciousness is a broad-band redistribution of EEG spectral power from higher to lower frequencies and the emergence of high-amplitude slow waves, a phenomenon known as ‘EEG slowing’^19,20. This phenomenon has been observed during the transition from wakefulness to deep sleep^21,22,23, during anesthetic-induced loss of consciousness (as reviewed in Brown et al., 2010²⁴) and in disorders of consciousness ^{25,26,27,28,29}. Moreover, regional increases of slow EEG activity were found to predict the absence of dream reports during sleep, suggesting that slowing may represent a useful index in cases of disconnected consciousness³⁰. Thus, studying regional slowing of cortical activity before and after hemispherotomy could provide useful information regarding the possibility of preserved consciousness in the disconnected cortex.

To date, the electrophysiological state of the disconnected cortex in behaviorally awake subjects remains unknown. Indeed, the sole EEG study of hemispherotomy conducted so far was performed under general anesthesia and could not resolve this issue³¹. Here, we fill this gap by studying regional EEG slowing during wakefulness in ten pediatric participants who had undergone a complete cortical disconnection, of either an entire hemisphere (complete hemispherotomy in 7 cases) or a large portion of it (temporo-parieto-occipital disconnection in 3 cases). We first evaluate narrow-band slowing, pre- and post-disconnection, in both the disconnected and its homologous contralateral cortex, by considering power within the Slow-Delta band (0.5-2 Hz). Further, to provide a comprehensive quantification of broad-band electrophysiological slowing, we estimated the spectral exponent, which reflects the 1/f-like decay of the Power Spectral Density (PSD) and thus the relative contribution of high and low frequencies. Notably, the spectral exponent has been proven as a reliable index of consciousness across various physiological^{32,33,34,35,36,37,38,39,40} pharmacological^{41,42,43,44,45} and pathological conditions^18,42,44, including states of disconnected consciousness. Crucially in the present context, the spectral exponent has been extensively validated in large pediatric samples to distinguish between wakefulness^46,47,48,49 and various sleep stages, including deep NREM sleep^35,46.

Patients receiving disconnective surgery

We retrospectively selected our participants starting from a large sample of pediatric patients referred from 2005 to 2020 to the Munari Center for Epilepsy Surgery of Niguarda Hospital in Milan. Ten participants were recruited on the basis of the following inclusion criteria: diagnosis of focal drug-resistant epilepsy with structural etiology on a malformative basis or as an outcome of perinatal hypoxic-ischemic insult; patient undergoing complete cortical disconnection of an hemisphere or of a large portion of it (at least one lobe); post-intervention outcome classified in class I according to the Engel Surgical Outcome Scale; availability of an EEG within 12 months prior to surgery and at least six months after surgery; age at pre-surgery EEG recording between 2 and 17 years; presence of at least 3 min of quiet wakefulness free from artifacts and interictal epileptiform discharges.

During the EEG the participants, lying on a bed or in the arms of the parent, did not perform any particular task, except alternating periods of dozens of seconds with eyes open and with eyes closed, following a constant protocol across participants. Since we were not interested in the neural basis of a specific content of consciousness, we considered these epochs jointly (as in Favaro et al., 2023⁴⁶). We excluded from the analysis all epochs in which the clinicians performed activation tests (intermittent light stimulation and hyperpnea). When necessary, younger children watched cartoons in order to improve compliance and reduce motion artifacts. EEG recordings were performed and used for research purposes after acquiring a written informed consent from the participants’ parents. The study was conducted in accordance with the Declaration of Helsinki and the institutional guidelines and the protocol was approved by the Ethics Committee of Niguarda Hospital, Milan, Italy (protocol number ID 348-24.06.2020)

Data acquisition and pre-processing

Each EEG recording was obtained by 19 scalp electrodes (cup in chlorinated silver) positioned according to the international 10–20 system and a 32-channel amplifier (Neurofax EEG-1200, Nihon Kohden Corporation). In acquisition, an additional electrode located between the Fp1 and Fp2 electrodes was used as a reference; the impedances of all the electrodes were kept below 5 kΩ for the entire duration of the recording. The signal was filtered during acquisition (high-pass filter 0.016 Hz; low-pass filter 300 Hz) and sampled at 500 Hz. The raw data was imported and analyzed with custom MATLAB code (MATLAB 9.7.0 R2019b, The MathWorks Inc., Natick, MA, USA). The signal was filtered with a 5th order Butterworth high-pass filter at 0.1 Hz (effectively retaining Slow-Delta frequency activity between 0.5 Hz and 2 Hz), with a 3rd order Butterworth low-pass filter at 60 Hz and with a 50 Hz Notch filter. Epochs characterized by artifacts and interictal epileptiform discharges (spikes, polyspikes, sharp waves, spike/polyspike-wave complexes) were visually selected and excluded by a trained pediatric neurologist expert in EEG (Supplementary Fig. 1). Due to the careful application of electrodes in our low-density set-up, recordings from all electrodes were deemed of sufficiently good quality. Electrodes were first re-referenced to the common average, to ease the topographical inspection of the components from a subsequent blind source separation.

Specifically, to minimize the influence of electromyographic (EMG) and electro-oculographic (EOG) activity, Independent Component Analysis was performed on a copy of the data, filtered with a narrower high-pass filter (5th order Butterworth at 0.5 Hz) to better separate artifactual from neurophysiological components. The different components of the signal were visually inspected on the basis of their time-series, topographies and PSD. Clearly identifiable components resembling EMG and EOG activity were marked for rejection. The unmixing weights (estimated from the data filtered with a 0.5 Hz high-pass) were applied to the original data (i.e. that filtered with a 0.1 Hz high-pass), and the retained components were back-projected to the scalp, thus effectively removing artifactual influences from the scalp electrodes signal⁵⁰.

Finally, to observe lateralized effects with high spatial specificity under our low-density EEG montage, bipolar derivations between all pairs of nearest neighboring electrodes (excluding the midline) were considered for subsequent analyses. The common average reference was used exclusively for ICA preprocessing and for topographic visual inspection and display (Fig. 1B, center).

Estimation of slow delta power and spectral exponent

We estimated narrow-band EEG slowing by means of Slow-Delta power and broad-band slowing by means of the spectral exponent. The PSD was estimated using Welch’s method, with a 3 second Hanning window and a 50% overlap, following linear detrending of the time-series in each periodogram. Only periodograms that were entirely free from artifacts and from epileptiform activity were included. The estimation of low frequency activity (above 0.5 Hz) was free from artifacts induced by the filter (0.1 Hz high-pass, as previously mentioned).

Slow-Delta power was estimated as the mean absolute PSD between 0.5 and 2 Hz, subsequently transformed by log10.

The spectral exponent indexes the slope of the overall PSD decay across frequencies, and thus characterizes the broad-band relative distribution between high and low frequencies (1/f-like). The neurophysiological 1/f-like distribution originates from aperiodic and quasi-periodic activity^51,52.

The exponent was estimated from a linear fit of the PSD under double logarithmic axes, after discarding frequency bins with peaks corresponding to periodic activity, i.e. bins with large positive residuals from a preliminary linear fit, and their contiguous bins with positive residuals (Supplementary Methods: fitting of the spectral exponent, Supplementary Methods: estimation of aperiodic activity while accounting for periodic activity, Supplementary Fig. 2 and Colombo et al., 2019⁴¹).

Here, we estimated the spectral exponent over the 0.5–20 Hz range, so as to include the Slow-Delta range, while excluding higher frequencies which are prone to muscular contamination in children. A similar fitting range, 1–20 Hz, previously used in Favaro et al., 2023⁴⁶, was also investigated in our patients’ sample for completeness (Supplementary Results: Results in the patient dataset of the Spectral Exponent, estimated between 1–20 Hz; Supplementary Fig. 3).

Subsequently, the median value of each feature (Slow-Delta power and spectral exponent) was taken across the lateralized bipolar derivations placed over the isolated cortex (15 derivations in the case of a hemispherotomy and 6 for a temporo-parieto-occipital disconnection) and the corresponding homologous bipolar derivations placed over the contralateral cortex.

Statistical comparisons: spatial inter-hemispheric differences, and temporal pre-post differences

We estimated narrow-band and broad-band EEG slowing, by means of Slow-Delta power and of the spectral exponent, before and after surgical disconnection, in both the disconnected and in the homologous contralateral cortex. For each of the two EEG features, we thereby assessed whether EEG slowing differed between homologous cortexes (spatial effect), between the recording sessions (pre vs post-surgery, temporal effect), and if it differently changed in the two cortexes following surgery (spatio-temporal interaction effect), by means of an ANOVA on a mixed effects model (Supplementary Methods: Mixed effects model).

Then, according to the ANOVA results, a set of contrasts across spatial and temporal conditions was performed by means of planned paired t-tests, for each EEG feature.

Specifically, we assessed a spatial inter-hemispheric contrast for each session (i.e., contrast of values between cortexes, in the pre- and in the post- surgery session; Fig. 2B, D), and a pre-post-surgery temporal contrast for each cortex (i.e. contrast of the pre-post difference against 0, in the disconnected and in the contralateral cortex, Fig. 2C, E). Further, we ascertained whether the pre-post difference was larger for the disconnected cortex (i.e., contrast of the pre-post difference between cortices; Fig. 2C, E); a finding which, in our specific design where measurements were paired over both time and space, would imply that the inter-hemispheric difference changed following surgical disconnection.

Spectral exponent: comparison to reference values during wake and sleep

To further establish whether spectral exponent values observed in patients with cortical disconnection were compatible with wakefulness or sleep, we included a reference dataset of spectral exponent values obtained from the EEG of a large sample of neurotypical pediatric participants during wakefulness and sleep⁴⁶. In this previous study, the EEG was obtained during a daytime nap opportunity in an outpatient clinic, in children and adolescents aged from 2 to 17 years, who showed no neurological abnormalities. Here, to build the spectral exponent reference dataset, we considered only the subset of participants who were able to fall asleep and reached all NREM sleep stages (up to N3), consisting of 44 participants (24 males, aged between 2–16 years, mean 6.798, s.d.=3.944).

To directly compare spectral exponent values across datasets, we aligned the preprocessing pipeline of the patients with cortical disconnection to that of the previously published study on the reference pediatric sample⁴⁶. Hence, in the patients’ dataset, we increased the Butterworth high-pass filter from 0.1 to 0.5 Hz, and consequently restricted the fitting range of the spectral exponent from 0.5–20 Hz to 1–20 Hz (i.e. with the lower bound set at a frequency where the filter effects were dissipated and the expected PSD decay was observed), according to the procedure used in the reference dataset⁴⁶. Even after restricting the frequency range under study, the findings on the spectral exponent were consistently replicated in the patient dataset (Supplementary Results: Results in the patient dataset of the Spectral Exponent, estimated between 1–20 Hz).

Finally, we adapted the virtual EEG reference of the pediatric reference dataset to that of the patient dataset, which required high spatial specificity. Hence, the spectral exponent of the reference dataset (previously estimated using average reference) was re-estimated over all lateralized nearest-neighbor bipolar derivations (15 per side, 30 total) and the median value was considered.

Patients and EEG recordings

We recruited 10 pediatric participants (four females). They had a diagnosis of focal drug-resistant epilepsy with structural etiology on a malformative basis (N = 4; pachygyria, polymicrogyria, focal cortical dysplasia) or as an outcome of perinatal hypoxic-ischemic insult (N = 6); patients underwent a complete disconnection of a large cortical portion: either an entire hemisphere (complete hemispherotomy, 7 patients), or the temporo-parieto-occipital region (3 patients). The baseline EEG was performed within 12 months before surgery (mean = 5.6, std = 4.67, range = [1, 12]) and at least six months after surgery (delay in months: mean = 19.8, std = 9.6, range = [6.96, 34.92]). The age at pre-surgery EEG spanned from 2 to 11 years (mean = 7.8, std = 2.9); in each recording we identified at least 3 min of quiet wakefulness free from artifacts and interictal epileptiform discharges (mean = 7.379, std = 3.489 range = [3.561, 12.815]). Demographic and clinical information for each patient are reported in Supplementary Table 3.

Surgical disconnection and EEG slowing

In this section we qualitatively describe the patterns observed in a representative subject, then consider the PSD and its underlying 1/f-like trend, aggregated across participants. Figure 1 illustrates the appearance of pronounced EEG slowing over the disconnected cortex following surgery, in an 11-year-old girl with a left porencephalic lesion secondary to perinatal stroke. The post-surgical MRI (Fig. 1A) highlights the disconnected cortex (in red) and the resection of all interhemispheric and subcortical connections to the left hemisphere. Short EEG segments recorded before (top) and after surgery (bottom) from both the disconnected cortex and contralateral homologous sites are shown in Fig. 1B. The visual inspection of the EEG time-series and of the topographies of low-frequency PSD (Slow-Delta power, 0.5-2 Hz) reveal high-amplitude (> 50 microvolts) slow waves emerging over the disconnected cortex post-surgery, associated with pronounced inter-hemispheric asymmetry. Further, the PSD profiles (Fig. 1C) indicate that the narrow band increase in the Slow-Delta band was accompanied by an overall redistribution of power from high to low frequencies, resulting in a steeper broad-band decay of the disconnected cortex following surgery. These observations were quantified by the Slow-Delta power and the spectral exponent (indexing the slope of the PSD decay over the 0.5–20 Hz range) over intra-hemispheric bipolar derivations (Fig. 1D).

These single-subject findings were consistently observed at the whole sample level. When comparing the seven patients with complete hemispherotomy to those who underwent temporo-parieto-occipital disconnection, we observed a highly similar pattern of localized increase in the degree of EEG slowing in the disconnected cortex (Supplementary Results: Comparing EEG slowing after cortical disconnection between patients with Hemispherotomy and those with Temporo-Parieto-Occipital disconnection). Results are thus reported across all 10 patients, irrespective of the extent of the surgery. As shown in Fig. 2, at the population level the disconnected cortex following surgery displayed increased Slow-Delta power and an overall PSD redistribution towards low frequencies and a steeper broad-band decay, as evident from the PSD and its 1/f-like fit, aggregated across patients. These observations were confirmed by statistical analysis, reported in the next section.

Statistical comparisons across hemispheres and pre- post- surgery

For each of the two EEG dependent variables (Slow Delta Power and Spectral Exponent, indexing narrow-band and broad-band slowing respectively), the ANOVA on the mixed effects model revealed a main effect of time (both p < 0.001), demonstrating a change due to surgical disconnection. Further, it also revealed a significant interaction effect between time and space (both p < 0.001), indicating that surgery differently affected the disconnected vs the contralateral cortex, and, similarly, that the inter-hemispheric asymmetry changed following surgery (for details see Supplementary Methods: Mixed effects models and Supplementary Table 1).

The specific effects of cortical disconnection on each EEG feature were explored through a series of planned pairwise contrasts, t-tests with 9 d.f. (Fig. 3, Supplementary Table 2).

Contrasting the pre-post difference between cortices, the disconnected cortex underwent a greater change following surgery as compared to the contralateral, for both the Slow-Delta power (T(9)= -3.146, p = 0.0118, greater change observed in 8/10 patients, Fig. 2C) and the spectral exponent (T = 6.464, p = 0.0001, greater change observed in 9/10 patients, Fig. 2E), indicating also that the inter-hemispheric asymmetry increased following surgery.

When assessing spatial inter-hemispheric contrasts (contralateral vs disconnected cortex), we found that Slow-Delta power was significantly larger in the pathological cortex than in the contralateral cortex, but only in the post-surgery session (Pre: T=-0.228, p = 0.8246; larger in 6/10 patients; Post: T = -4.245, p = 0.0022, larger in 9/10 patients, Fig. 2B); the spectral exponent was significantly more negative in the disconnected cortex than in the contralateral, both pre-surgery (T = 2.395, p = 0.0402; more negative in 7/10 patients) and, especially, post-surgery (T = 7.118, p < 0.0001, more negative in 10/10 patients, Fig. 2D).

Assessing temporal contrasts (Post-Pre surgery), the Slow-Delta power was larger following surgery, significantly only in the disconnected cortex (contralateral: T= -1.462, p = 0.1777, larger in 9/10 patients; disconnected: T= -3.215, p = 0.0106; larger in 10/10 patients, Fig. 2C); similarly, the spectral exponent was significantly more negative following surgery, only in the disconnected cortex (contralateral: T = 1.786, p = 0.1077, more negative in 5/10 patients; disconnected: T = 6.455, p = 0.0001, more negative in 10/10 patients, Fig. 2E).

Overall, before surgery we found no evidence for interhemispheric difference in narrow-band Slow-Delta power and only a small, albeit significant, inter-hemispheric asymmetry in the degree of broad-band slowing. Instead, after surgery, we observed a marked inter-hemispheric asymmetry characterized by prominent increase in both the degree of EEG narrow-band and, especially, of broad-band slowing over the disconnected cortex. The spectral exponent revealed larger effect sizes than delta-power, thus potentially providing a more sensitive index of the neurophysiological consequences of hemispherotomy (Supplementary Results: Comparing narrow-band vs broad-band slowing in cortical isolation

and Supplementary Fig. 4).

Comparison of spectral exponent to a reference sample during wake and sleep

Finally, the spectral exponent values assessed in patients before and after cortical disconnection were compared to the values assessed in a reference pediatric sample during wakefulness and NREM sleep (Fig. 3). Qualitatively, the typical values (i.e., those of the central half of the sample, within the interquartile range) of the contralateral cortex overlapped with the typical values observed during wakefulness and N1 in the reference sample. The typical values of the disconnected cortex before surgery overlapped with the typical wakefulness and N1 values of the reference sample; after surgery, the typical values of the disconnected cortex overlapped with the typical N2 and N3 values of the reference sample. Notably, 9/10 values from the disconnected cortex did not overlap with the wakefulness distribution; and the 10th case was below all but one of the values from the wakefulness distribution.

Further, we compared the mean spectral exponent values of each condition across the two datasets through bootstrap resampling. Specifically, we considered the bootstrapped distribution of the pairwise differences between the mean values of the pediatric reference participants (either during wakefulness, N1, N2, or N3 sleep) and the mean values of the patients (in the disconnected or contralateral cortex, before or after surgery) (Supplementary Results: Comparison of spectral exponent with the reference distribution of sleep and wake; Supplementary Fig. 5). Accordingly, the homologous contralateral cortex, both before and after surgery, showed values distinct from the NREM sleep distributions and closest to the wakefulness distribution, albeit on the lower end. The disconnected cortex before surgery showed values closest to the N1 sleep distribution; crucially, after surgery the disconnected cortex showed values distinct from the wakefulness distribution and intermediate between those of the N2 and N3 sleep distributions.

This is the first exploration of the electrophysiological changes occurring in awake human participants before and after surgical disconnection of extensive regions of the cerebral cortex, including a complete hemisphere. Unlike the contralateral homologous cortex, after surgical disconnection the isolated cortex entered a state dominated by large-amplitude slow wave activity (below 2 Hz), and an overall redistribution of the PSD from high to low frequencies. Such broad-band neurophysiological slowing, as indexed by the spectral exponent, was similar to that previously found in the unconscious states seen in deep sleep, anesthesia and brain-injury.
Upon direct comparison of the spectral exponent with that of a reference pediatric sample previously recorded during both wakefulness and sleep, we found that the disconnected cortex exhibited values consistent with deep NREM sleep (N2 and N3), while the contralateral cortex retained values indicative of wakefulness.

Mechanisms of EEG slowing in cortical disconnection

Following surgery, a pronounced increase in Slow-Delta power (0.5 - 2 Hz) appeared over the disconnected cortex. The contralateral cortex, retaining intact subcortical inputs, did not display any significant increase in the degree of EEG slowing following surgery, despite the surgical resection of all its inter-hemispheric connections. This finding, in line with neurophysiological evidence in callosotomy^53,54, suggests that the primary cause of EEG slowing observed in the disconnected cortex lies in the deafferentation from subcortical, rather than cortical, inputs.

Notably, despite the disconnected patients being awake during EEG recordings, the electrophysiology of the disconnected cortex showed some striking similarities with the electrophysiology of sleep. During NREM sleep, EEG slow waves emerge throughout the cerebral cortex as a consequence of decreased levels of activating neuromodulation from brainstem activating systems⁵⁵. A similar pattern can be observed in pathological conditions following lesions or compressions in the brainstem and midbrain ascending activating systems^56,57. More generally, slow waves can be found in various experimental models in which cortical circuits are disconnected from subcortical inputs, such as in isolated cortical gyri (i.e., cortical slabs)⁵⁸, in cortical slices in vitro⁵⁹, as well as after thalamic inactivation⁶⁰. As such, cortical slow waves are considered the elemental intrinsic activity pattern of cortical networks⁶¹. The present study in hemispherotomy provides further evidence for this account of cortical slow waves, showing that the deprivation of all ascending influences leads the cortex to a state characterized by sleep-like slow waves, in an alert and behaviorally awake subject.

At the same time, the overall EEG state of the isolated cortex clearly differed from natural sleep, during which cortical slow oscillations are enriched and orchestrated by subcortical, especially thalamic, dynamics. Specifically, after hemispherotomy slow EEG oscillations were not accompanied by visible spindle activity, as confirmed by the absence of a sigma peak in the PSD (Supplementary Fig. 2). This is not surprising, as spindles are generated within the thalamus and are typically synchronized to slow waves through cortico-thalamo-cortical loops⁶². Another difference was that the EEG background activity observed in the disconnected cortex, despite being dominated by high-amplitude slow waves, did not fully attain the PSD decay typical of N3 sleep in healthy intact brains. This difference may also be attributed to thalamic disconnection. Indeed, the integrity of cortico-thalamo-cortical loops is known to play an important role in synchronizing EEG slow oscillations⁶³, likely explaining the steeper decay of the PSD distribution in physiological sleep as compared to cortical disconnection.

An island of sleep

Inferring consciousness in isolated human cortical hemispheres represents a unique challenge^7,8. For example, the pathological cortex differs from lab-grown cortical organoids, because it has developed as part of a human brain in contact with the external world, and thus some arguments against the possibility of consciousness in cortical organoids would not apply to hemispherotomy⁶⁴. A vast literature on acquired brain-injury^65,66,67, hemispherectomy^68,69 and split-brain patients^70,71 suggests that one hemisphere is sufficient to support consciousness. However, hemispherotomy also differs from split-brain surgery, where the two hemispheres are disconnected from each other but retain all subcortical afferents and their ability to interact with the external world^70,71. Hence, whether the isolated cortex represents a distinct locus of consciousness from that supported by the contralateral hemisphere remains an open question with important scientific and ethical implications.

The possibility of an island of inaccessible awareness within the skull is suggested by recent fMRI studies indicating preserved neural networks lateralized within the isolated hemisphere and specifically by the integrity of the default mode network (DMN)^{9, 10} This finding is particularly intriguing because, in healthy awake participants, this network is typically activated during self-directed mentation, such as daydreaming and reflection. Yet, resting state networks, including the DMN, can be disrupted in conscious psychedelic states⁷² and, crucially, preserved in unconscious states such as sleep and anesthesia^11,12,13,14, challenging the interpretation of the presence of fMRI network connectivity as evidence for preserved consciousness. In this context, it is crucial to better assess the nature of neuronal activity within the disconnected cortex, beyond fMRI connectivity patterns.

The present findings of prominent slow waves and steep spectral decay strongly suggest that the disconnected cortex rests in an electrophysiological state that is not compatible with the presence of an island of awareness. This conclusion is supported by different lines of converging evidence. First, EEG slow waves are canonically associated with unconscious conditions encompassing NREM sleep^73,74,75, general anesthesia^76,77,78,79 and the vegetative state^{17,18,26,27,29,80}. As shown by animal and human studies, EEG slow waves reflect the tendency of cortical neurons to become silent after an initial activation, a mechanism that prevents cortical circuits from engaging in the complex network interactions normally observed in conscious states^75,80,81,82.

Second, recent studies in healthy humans have shown that increased low frequency power and decreased high frequency power are predictive of the absence of dream reports during sleep⁸³. Therefore, the present finding of a redistribution of power from high to low frequencies detected by the spectral exponent (Fig. 1 and Fig. 2) can be interpreted as an indication of the reduced likelihood of dream-like experiences in the isolated cortex.

Third, the spectral exponent has been previously tested as an index of consciousness in large samples of healthy controls and patients, including challenging conditions of sensorimotor disconnection and behavioral unresponsiveness. In good agreement with perturbational measures of brain complexity (the Perturbational Complexity Index⁸⁴), the spectral exponent robustly discriminated between conscious and unconscious states, across wakefulness, sleep, physiological or pharmacological dreams, anesthesia induced by different compounds, and severe brain injury^18,41,43,45. Furthemore, unlike other clinical brain-based markers of consciousness, the spectral exponent has been validated across the sleep-wake cycle in a reference pediatric sample⁴⁶, spanning the typical age range of patients undergoing cortical disconnection.

Capitalizing on this extensive validation, we here employed the spectral exponent as a quantitative index to extrapolate the global state of consciousness in the disconnected human cortex. After surgery, the spectral exponent of the isolated cortex plunged to values typically associated with loss of consciousness in benchmark conditions and departing from the wakefulness distribution of the pediatric population in 9/10 cases (Fig. 3)—compatible with N2-N3 sleep, as confirmed by bootstrap analysis (see Supplementary Figure 5). That the spectral exponent attained such negative values is remarkable, if one considers that the isolated cortex lacks the synchronizing drive of slow rhythms provided by cortico-thalamo-cortical loops. On the other hand, this quantification of the overall spectral profile is consistent with the visual observation that slow waves dominated the EEG background activity of the isolated cortex.

Detecting an electrophysiological pattern similar to that observed in reference physiological, pharmacological and pathological conditions in which the global state of consciousness is abolished or greatly reduced, mitigates the concerns that hemispherotomy may result in cortical islands of awareness. More generally, the present study builds on previous fMRI explorations^9,10, to show how different brain-based tests can be used in an iterative process to evaluate questions about the possibility of consciousness in challenging conditions⁸⁵. In this vein, the present electrophysiological investigation motivates reinterpreting previous evidence and suggests that the partial integrity of resting-state networks may not suffice for consciousness in the isolated cortex.

Limitations and future directions

A first general limitation of the present study is that any inference regarding the presence/absence of consciousness purely based on physical properties of the brain (including spontaneous EEG dynamics) in neural structures that are not behaviorally accessible must be made cautiously. In particular, the possibility that some form of consciousness might be retained even in the presence of large slow waves in the spontaneous scalp EEG should be considered, especially when it comes to pathological conditions⁸⁶and needs to be further explored. High density recordings of the disconnected hemisphere may be useful to confirm the pervasiveness of sleep-like slow waves beyond the coarse spatial resolution offered by standard EEG recordings.

Perturbation approaches, combining transcranial magnetic stimulation and EEG, have shown high sensitivity in detecting consciousness also in the presence of high delta power^87,88and may be used to directly assess the capacity of the isolated cortex to sustain complex patterns of causal interactions, similar to those typically found in conscious subjects⁸⁹. Also, in the cases where olfactory pathways to the isolated cortex are preserved, it will be important to assess whether high-level processing, such as the discrimination between familiar and unfamiliar odors⁸, is retained.

Another interesting element to consider is that, even before surgery, the pathological cortex showed slightly lower mean values than its contralateral counterpart, with a distribution centered between the wakefulness and the N1 sleep distributions of the reference pediatric population (Fig. 3). Such relative slowing in baseline conditions may be contributed by interictal epileptiform discharges⁹⁰—although these graphoelements were carefully excluded by a trained neurologist during data pre-processing (see Supplementary Figure 1)—as well as by the effects of focal cortical malformations and gliosis ^91,92,93. In this perspective, the role of the pathological hemisphere and the nature of its contribution to consciousness before surgery remains an open question.

Finally, the present study prompts further investigations of the relationships between EEG and fMRI activity patterns. Although the functional organization of macroscale resting-state networks (including the DMN) is retained in the isolated hemisphere, previous works also found increased intrahemispheric connectivity and loss of the normal anticorrelations between networks ^9, ¹⁰. Recent rodent studies⁹⁴ combining electrophysiological and fMRI recordings point to a relationship between the emergence and propagation of EEG slow waves and functional network abnormalities, including DMN hyperconnectivity. Similarly, clinically relevant alterations involving both EEG slowing and functional network abnormalities—including loss of normal anticorrelations, reduced interhemispheric connectivity along with increased intrahemispheric connectivity--can be found in patients with stroke⁹⁵. In this perspective, hemispherotomy and the coexistence of unihemispheric sleep-like and wakefulness activity patterns within the same brain may offer a unique model to better elucidate the relationships between electrophysiological dynamics and fMRI functional networks.

Cortical isolation of the left hemisphere in a 11 y.o. representative patient. (A) The anatomical MRI 6 months after surgery shows that all interhemispheric and subcortical connections were resected (sagittal, axial and coronal panels are shown from left to right). The cortical surface is highlighted in red for the disconnected cortex (in blue for the contralateral). (B) EEG recordings were performed 8 months before surgery (at 11 y.o.), and 12 months after. Short EEG segments (6 s) from the isolated cortex (F7-F3 and T5-P3 bipolar derivations, left panel) and the contralateral homologue (F8-F4 and T6-P4, right panel) are shown, pre-surgery (above) and post-surgery (below). In the middle panel, the topographies of Slow-Delta power, indexing the power spectral density (PSD) of low-frequency activity (0.5-2 Hz, hereby estimated under average reference for display purposes), show a marked inter-hemispheric asymmetry post-surgery. (C) The EEG PSD, averaged by geometric mean across intrahemispheric bipolar derivations, reveal an increase of slow frequency activity (< 2 Hz) and a broad-band steepening of the PSD over the disconnected hemisphere post-surgery, which is evidenced in the inset graph by the the linear fit of the PSD under double logarithmic axes (power-law fit). (D) For all the neighboring intrahemispheric bipolar derivations, we display the values of Slow-Delta power (0.5-2 Hz), and of the spectral exponent, indexing the slope of the PSD decay (0.5-20 Hz).

Authorship

Conceptualization: M.A.C., J.F., E.M., Si.Sa., T.B., A.K.S., M.M. Methodology: M.A.C., E.M. A.P., Si.Sa. Software: M.A.C., E.M. Validation: M.A.C., J.F. Formal analyses: M.A.C., E.M. Investigation: F.M.Z., I.S., P.D., L.C. Resources: A.P., F.M.Z., I.S., P.D., L.C. Data curation: J.F., A.P., F.M.Z., I.S. Writing - Original Draft: M.A.C., J.F. Writing - Review & Editing: M.A.C., J.F., E.M., A.P., St.Sa., I.T., Si.Sa., T.B., A.K.S., M.M. Visualization: M.A.C., J.F., E.M., A.P. Supervision: St.Sa., I.T., Si.Sa., T.B., A.K.S., M.M. Project administration and Funding acquisition: M.M.

Data availability

Data will be made available on request.

Acknowledgements

Thanks to all the children and their parents who made this work possible.

Jacopo Favaro thanks the TMO group for inspiring discussions and the constant support.

A.S. is supported by European Research Council Advanced Investigator Grant CONSCIOUS, grant number 101019254. T.B., A.S and M.M. acknowledge support provided by the Brain, Mind, and Consciousness Program of the Canadian Institute for Advanced Research (CIFAR).

F.M.Z. is supported by the Department of Philosophy “P. Martinetti” of the University of Milan under the Project “Departments of Excellence 2023-2027” awarded by the Ministry of University and Research (MUR)
Competing interests

Marcello Massimini is co-founder and shareholder of Intrinsic Powers, a spin-off of the University of Milan; Simone Sarasso is advisor of the same company. All other authors do not have any conflict of interest related to the present submission.

We prepare this manuscript in accordance with the STROBE checklist for observational studies.

Griessenauer CJ, Salam S, Hendrix P, et al. Hemispherectomy for treatment of refractory epilepsy in the pediatric age group: A systematic review. J Neurosurg Pediatr. 2015;15(1):34-44. doi:10.3171/2014.10.PEDS14155
Marras CE, Granata T, Franzini A, et al. Hemispherotomy and functional hemispherectomy: Indications and outcome. Epilepsy Res. 2010;89(1):104-112. doi:10.1016/j.eplepsyres.2009.09.006
de Ribaupierre S, Delalande O. Hemispherotomy and other disconnective techniques. Neurosurg Focus. 2008;25(3). doi:10.3171/FOC/2008/25/9/E14
Delalande O, Bulteau C, Dellatolas G, et al. Vertical parasagittal hemispherotomy: Surgical procedures and clinical long-term outcomes in a population of 83 children. Neurosurgery. 2007;60(2 SUPPL.1). doi:10.1227/01.NEU.0000249246.48299.12
de Palma L, Pietrafusa N, Gozzo F, et al. Outcome after hemispherotomy in patients with intractable epilepsy: Comparison of techniques in the Italian experience. Epilepsy and Behavior. 2019;93:22-28. doi:10.1016/j.yebeh.2019.01.006
Devlin AM, Cross JH, Harkness W, et al. Clinical outcomes of hemispherectomy for epilepsy in childhood and adolescence. Brain. 2003;126(3):556-566. doi:10.1093/brain/awg052
Bayne T, Seth AK, Massimini M. Are There Islands of Awareness? Trends Neurosci. 2020;43(1):6-16. doi:10.1016/j.tins.2019.11.003
Gauvry C, Rüber T. Extrapolating Consciousness in Isolated Hemispheres. Hemispherotomy as a New Challenge. Revue Metaphysique. 2024; E1:47-66.
Blauwblomme T, Demertzi A, Tacchela JM, et al. Complete hemispherotomy leads to lateralized functional organization and lower level of consciousness in the isolated hemisphere. Epilepsia Open. 2020;5(4):537-549. doi:10.1002/epi4.12433
Rüber T, Bauer T, Gauvry C, et al. Intact functional brain networks in the isolated hemisphere of people after hemispherotomy, 12 April 2024, PREPRINT (Version 1) available at Research Square [https://doi.org/10.21203/rs.3.rs-4138881/v1]
Boly M, Phillips C, Tshibanda L, et al. Intrinsic brain activity in altered states of consciousness: How conscious is the default mode of brain function? In: Annals of the New York Academy of Sciences. Vol 1129. Blackwell Publishing Inc.; 2008:119-129. doi:10.1196/annals.1417.015
Horovitz SG, Fukunaga M, De Zwart JA, et al. Low frequency BOLD fluctuations during resting wakefulness and light sleep: A simultaneous EEG-fMRI study. Hum Brain Mapp. 2008;29(6):671-682. doi:10.1002/hbm.20428
Vincent JL, Patel GH, Fox MD, et al. Intrinsic functional architecture in the anaesthetized monkey brain. Nature. 2007;447(7140):83-86. doi:10.1038/nature05758
Akeju O, Loggia ML, Catana C, et al. Disruption of thalamic functional connectivity is a neural correlate of dexmedetomidine-induced unconsciousness. Elife. 2014;3(November):1-23. doi:10.7554/eLife.04499
Bender A, Jox RJ, Grill E, Straube A, Lulé D. Persistent vegetative state and minimally conscious state-A systematic review and metaanalysis of diagnostic procedures. Dtsch Arztebl Int. 2015;112(14):235-242. doi:10.3238/arztebl.2015.0235
Brenner RP. The interpretation of the EEG in stupor and coma. Neurologist. 2005;11(5):271-284. doi:10.1097/01.nrl.0000178756.44055.f6
Comanducci A, Boly M, Claassen J, et al. Clinical and advanced neurophysiology in the prognostic and diagnostic evaluation of disorders of consciousness: review of an IFCN-endorsed expert group. Clinical Neurophysiology. 2020;131(11):2736-2765. doi:10.1016/j.clinph.2020.07.015
Colombo MA, Comanducci A, Casarotto S, et al. Beyond alpha power: EEG spatial and spectral gradients robustly stratify disorders of consciousness. Cerebral Cortex. 2023;33(11):7193-7210. doi:10.1093/cercor/bhad031
Oken BS, Salinsky MC, Elsas SM. Vigilance, alertness, or sustained attention: physiological basis and measurement. Clinical Neurophysiology. 2006;117(9):1885-1901. doi:10.1016/j.clinph.2006.01.017
Steyn-Ross ML, Steyn-Ross DA, Sleigh JW. Modelling general anaesthesia as a first-order phase transition in the cortex. In: Progress in Biophysics and Molecular Biology. Vol 85. ; 2004:369-385. doi:10.1016/j.pbiomolbio.2004.02.001
Merica H, Blois R. Relationship between the time courses of power in the frequency bands of human sleep EEG. Neurophysiol Clin. 1997;27(2):116-128. doi:10.1016/S0987-7053(97)85664-X
De Gennaro L, Ferrara M, Bertini M. The boundary between wakefulness and sleep: quantitative electroencephalographic changes during the sleep onset period. Neuroscience. 2001;107(1):1-11. doi:10.1016/s0306-4522(01)00309-8
Ogilvie RD. The process of falling asleep. Sleep Med Rev. 2001;5(3):247-270. doi:10.1053/smrv.2001.0145
Brown EN, Lydic R, Schiff ND. General Anesthesia, Sleep, and Coma. New England Journal of Medicine. 2010;363(27):2638-2650. doi:10.1056/nejmra0808281
Chennu S, Finoia P, Kamau E, et al. Spectral Signatures of Reorganised Brain Networks in Disorders of Consciousness. PLoS Comput Biol. 2014;10(10). doi:10.1371/journal.pcbi.1003887
Estraneo A, Loreto V, Guarino I, et al. Standard EEG in diagnostic process of prolonged disorders of consciousness. Clinical Neurophysiology. 2016;127(6):2379-2385. doi:10.1016/j.clinph.2016.03.021
Schiff ND. Mesocircuit mechanisms underlying recovery of consciousness following severe brain injuries: Model and predictions. In: Brain Function and Responsiveness in Disorders of Consciousness. Springer International Publishing; 2016:195-204. doi:10.1007/978-3-319-21425-2_15
Sitt JD, King JR, El Karoui I, et al. Large scale screening of neural signatures of consciousness in patients in a vegetative or minimally conscious state. Brain. 2014;137(8):2258-2270. doi:10.1093/brain/awu141
Lechinger J, Bothe K, Pichler G, et al. CRS-R score in disorders of consciousness is strongly related to spectral EEG at rest. J Neurol. 2013;260(9):2348-2356. doi:10.1007/s00415-013-6982-3
Siclari F, Tononi G. Local aspects of sleep and wakefulness. Curr Opin Neurobiol. 2017;44:222-227. doi:10.1016/j.conb.2017.05.008
Hawasli AH, Chacko R, Szrama NP, et al. Electrophysiological sequelae of hemispherotomy in ipsilateral human cortex. Front Hum Neurosci. 2017;11. doi:10.3389/fnhum.2017.00149
Miskovic V, MacDonald KJ, Rhodes LJ, Cote KA. Changes in EEG multiscale entropy and power-law frequency scaling during the human sleep cycle. Hum Brain Mapp. 2019;40(2):538-551. doi:10.1002/hbm.24393
Bódizs R, Szalárdy O, Horváth C, et al. A set of composite, non-redundant EEG measures of NREM sleep based on the power law scaling of the Fourier spectrum. Sci Rep. 2021;11(1). doi:10.1038/s41598-021-81230-7
Bódizs R, Schneider B, Ujma PP, Horváth CG, Dresler M, Rosenblum Y. Fundamentals of sleep regulation: Model and benchmark values for fractal and oscillatory neurodynamics. Prog Neurobiol. 2024;234. doi:10.1016/j.pneurobio.2024.102589
Schneider B, Szalárdy O, Ujma PP, et al. Scale-free and oscillatory spectral measures of sleep stages in humans. Front Neuroinform. 2022;16. doi:10.3389/fninf.2022.989262
Horváth C, Szalárdy O, Ujma PP, et al. Overnight dynamics in scale-free and oscillatory spectral parameters of NREM sleep EEG. Sci Rep. 2022;12(1). doi:10.1038/s41598-022-23033-y
Zilio F, Gomez-Pilar J, Cao S, et al. Are intrinsic neural timescales related to sensory processing? Evidence from abnormal behavioral states. Neuroimage. 2021;226. doi:10.1016/j.neuroimage.2020.117579
Alnes SL, Bächlin LZM, Schindler K, Tzovara A. Neural complexity and the spectral slope characterise auditory processing in wakefulness and sleep. European Journal of Neuroscience. Published online March 1, 2023. doi:10.1111/ejn.16203
Shen Y, Olbrich E, Achermann P, Meier PF. Dimensional complexity and spectral properties of the human sleep EEG. doi:10.1016/S1
Rosenblum Y, Esfahani MJ, Adelhöfer N, et al. Fractal cycles of sleep: a new aperiodic activity-based definition of sleep cycles. bioRxiv. Published online January 1, 2024:2023.07.04.547323. doi:10.1101/2023.07.04.547323
Colombo MA, Napolitani M, Boly M, et al. The spectral exponent of the resting EEG indexes the presence of consciousness during unresponsiveness induced by propofol, xenon, and ketamine. Neuroimage. 2019;189:631-644. doi:10.1016/j.neuroimage.2019.01.024
Maschke C, Duclos C, Owen AM, Jerbi K, Blain-Moraes S. Aperiodic brain activity and response to anesthesia vary in disorders of consciousness. Neuroimage. 2023;275. doi:10.1016/j.neuroimage.2023.120154
Maschke C, O'Byrne J, Colombo MA, et al. Criticality of resting-state EEG predicts perturbational complexity and level of consciousness during anesthesia. Preprint. bioRxiv. 2023;2023.10.26.564247. Published 2023 Oct 31. doi:10.1101/2023.10.26.564247
Zilio F, Gomez-Pilar J, Chaudhary U, et al. Altered brain dynamics index levels of arousal in complete locked-in syndrome. Commun Biol. 2023;6(1). doi:10.1038/s42003-023-05109-1
Casey CP, Tanabe S, Farahbakhsh ZZ, et al. Evaluation of putative signatures of consciousness using specific definitions of responsiveness, connectedness, and consciousness. Br J Anaesth. Published online October 30, 2023. doi:10.1016/j.bja.2023.09.031
Favaro J, Colombo MA, Mikulan E, et al. The maturation of aperiodic EEG activity across development reveals a progressive differentiation of wakefulness from sleep. Neuroimage. 2023;277. doi:10.1016/j.neuroimage.2023.120264
Cellier D, Riddle J, Petersen I, Hwang K. The development of theta and alpha neural oscillations from ages 3 to 24 years. Dev Cogn Neurosci. 2021;50. doi:10.1016/j.dcn.2021.100969
McSweeney M, Morales S, Valadez EA, et al. Age-related trends in aperiodic EEG activity and alpha oscillations during early- to middle-childhood. Neuroimage. 2023;269. doi:10.1016/j.neuroimage.2023.119925
Schaworonkow N, Voytek B. Longitudinal changes in aperiodic and periodic activity in electrophysiological recordings in the first seven months of life. Dev Cogn Neurosci. 2021;47. doi:10.1016/j.dcn.2020.100895
Winkler I, Debener S, Müller KR, Tangermann M. On the influence of high-pass filtering on ICA-based artifact reduction in EEG-ERP. Annu Int Conf IEEE Eng Med Biol Soc. 2015;2015:4101-4105. doi:10.1109/EMBC.2015.731929
Palva S, Palva JM. Roles of Brain Criticality and Multiscale Oscillations in Temporal Predictions for Sensorimotor Processing. Trends Neurosci. 2018;41(10):729-743. doi:10.1016/J.TINS.2018.08.008
Gerster M, Waterstraat G, Litvak V, et al. Separating Neural Oscillations from Aperiodic 1/f Activity: Challenges and Recommendations. Neuroinformatics. 2022;20(4):991-1012. doi:10.1007/s12021-022-09581-8
Pizoli CE, Shah MN, Snyder AZ, et al. Resting-state activity in development and maintenance of normal brain function. Proc Natl Acad Sci U S A. 2011;108(28):11638-11643. doi:10.1073/pnas.1109144108
Quattrini A, Papo I, Cesarano R, et al. EEG Patterns after callosotomy. J Neurosurg Sci. 1997;41(1):85-92.
McCormick DA, Williamson A. Convergence and divergence of neurotransmitter action in human cerebral cortex. Proc Natl Acad Sci U S A. 1989;86(20):8098-8102. doi:10.1073/pnas.86.20.8098
Lee SH, Dan Y. Neuromodulation of Brain States. Neuron. 2012;76(1):209-222. doi:10.1016/j.neuron.2012.09.012
McCormick DA. Neurotransmitter actions in the thalamus and cerebral cortex and their role in neuromodulation of thalamocortical activity. Prog Neurobiol. 1992;39(4):337-388. doi:10.1016/0301-0082(92)90012-4
Timofeev I, Grenier F, Bazhenov M, Sejnowski TJ, Steriade M. Origin of slow cortical oscillations in deafferented cortical slabs. Cereb Cortex. 2000;10(12):1185-1199. doi:10.1093/cercor/10.12.1185
Capone C, Rebollo B, Munoz A, et al. Slow waves in cortical slices: How spontaneous activity is shaped by laminar structure. Cerebral Cortex. 2019;29(1):319-335. doi:10.1093/cercor/bhx326
Lemieux M, Chen JY, Lonjers P, Bazhenov M, Timofeev I. The impact of cortical deafferentation on the neocortical slow oscillation. Journal of Neuroscience. 2014;34(16):5689-5703. doi:10.1523/JNEUROSCI.1156-13.2014
Sanchez-Vives M V., Massimini M, Mattia M. Shaping the Default Activity Pattern of the Cortical Network. Neuron. 2017;94(5):993-1001. doi:10.1016/j.neuron.2017.05.015
Steriade M. Grouping of brain rhythms in corticothalamic systems. Neuroscience. 2006;137(4):1087-1106. doi:10.1016/j.neuroscience.2005.10.029
Crunelli V, Hughes SW. The slow (1 Hz) rhythm of non-REM sleep: A dialogue between three cardinal oscillators. Nat Neurosci. 2010;13(1):9-17. doi:10.1038/nn.2445
Croxford J, Bayne T. The Case Against Organoid Consciousness. Neuroethics. 2024;17(1). doi:10.1007/s12152-024-09548-3
Gschaidmeier A, Heimgärtner M, Schnaufer L, et al. Non-verbal Intelligence in Unilateral Perinatal Stroke Patients With and Without Epilepsies. Front Pediatr. 2021;9. doi:10.3389/fped.2021.660096
Asaridou SS, Demir-Lira E, Goldin-Meadow S, Levine SC, Small SL. Language development and brain reorganization in a child born without the left hemisphere. Cortex. 2020;127:290-312. doi:10.1016/j.cortex.2020.02.006
Sebastianelli L, Versace V, Nothdurfter W, et al. Functional reorganization after hemispherectomy in humans and animal models: What can we learn about the brain’s resilience to extensive unilateral lesions? Brain Res Bull. 2017;131:156-167. doi:10.1016/j.brainresbull.2017.04.005
Lew SM. Hemispherectomy in the treatment of seizures: a review. Transl Pediatr. 2014;3(3):208-217. doi:10.3978/J.ISSN.2224-4336.2014.04.01
Fisher PM, Albrechtsen SS, Nersesjan V, Amiri M, Kondziella D. Case Report: Resting-State Brain-Networks After Near-Complete Hemispherectomy in Adulthood. Front Neurol. 2022;13. doi:10.3389/FNEUR.2022.885115
Schechter E, Bayne T. Consciousness after split-brain surgery: The recent challenge to the classical picture. Neuropsychologia. 2021;160. doi:10.1016/j.neuropsychologia.2021.107987
F de Haan EH, Corballis PM, Hillyard SA, et al. Split-Brain: What We Know Now and Why This is Important for Understanding Consciousness. Brain. 2018;46(6):15. doi:10.1007/s11065-020-09439-3/Published
Gattuso JJ, Perkins D, Ruffell S, et al. Default Mode Network Modulation by Psychedelics: A Systematic Review. Int J Neuropsychopharmacol. 2023;26(3):155-188. doi:10.1093/IJNP/PYAC074
Jouvet M. Neurophysiology of the states of sleep. Physiol. Rev., 47 (1967), pp. 117-177
Nir Y, Staba RJ, Andrillon T, et al. Regional Slow Waves and Spindles in Human Sleep. Neuron. 2011;70(1):153-169. doi:10.1016/j.neuron.2011.02.043
Pigorini A, Sarasso S, Proserpio P, et al. Bistability breaks-off deterministic responses to intracortical stimulation during non-REM sleep. Neuroimage. 2015;112:105-113. doi:10.1016/j.neuroimage.2015.02.056
Purdon PL, Sampson A, Pavone KJ, Brown EN. Clinical Electroencephalography for Anesthesiologists: Part I: Background and Basic Signatures. Anesthesiology. 2015;123(4):937-960. doi:10.1097/ALN.0000000000000841
Murphy M, Bruno MA, Riedner BA, et al. Propofol anesthesia and sleep: a high-density EEG study. Sleep. 2011;34(3):283-91A. Published 2011 Mar 1. doi:10.1093/sleep/34.3.283
Lewis LD, Weiner VS, Mukamel EA, et al. Rapid fragmentation of neuronal networks at the onset of propofol-induced unconsciousness. Proc Natl Acad Sci U S A. 2012;109(49). doi:10.1073/pnas.1210907109
Sarasso S, Boly M, Napolitani M, et al. Consciousness and complexity during unresponsiveness induced by propofol, xenon, and ketamine. Current Biology. 2015;25(23):3099-3105. doi:10.1016/j.cub.2015.10.014
Rosanova M, Fecchio M, Casarotto S, et al. Sleep-like cortical OFF-periods disrupt causality and complexity in the brain of unresponsive wakefulness syndrome patients. Nat Commun. 2018;9(1). doi:10.1038/s41467-018-06871-1
D’Andola M, Rebollo B, Casali AG, et al. Bistability, causality, and complexity in cortical networks: An in vitro perturbational study. Cerebral Cortex. 2018;28(7):2233-2242. doi:10.1093/cercor/bhx122
Arena A, Comolatti R, Thon S, Casali AG, Storm JF. General anesthesia disrupts complex cortical dynamics in response to intracranial electrical stimulation in rats. eNeuro. 2021;8(4). doi:10.1523/ENEURO.0343-20.2021
Siclari F, Baird B, Perogamvros L, et al. The neural correlates of dreaming. Nat Neurosci. 2017;20(6):872-878. doi:10.1038/nn.4545
Casali AG, Gosseries O, Rosanova M, et al. A theoretically based index of consciousness independent of sensory processing and behavior. Sci Transl Med. 2013;5(198). doi:10.1126/scitranslmed.3006294
Bayne T, Seth AK, Massimini M, et al. Tests for consciousness in humans and beyond. Trends Cogn Sci. Published online 2024. doi:10.1016/j.tics.2024.01.010
Frohlich J, Toker D, Monti MM. Consciousness among delta waves: A paradox? Brain. 2021;144(8):2257-2277. doi:10.1093/brain/awab095
Comanducci A, Casarotto S, Rosanova M, et al. Unconsciousness or unresponsiveness in akinetic mutism? Insights from a multimodal longitudinal exploration. European Journal of Neuroscience. Published online March 1, 2023. doi:10.1111/ejn.15994
Casarotto S, Hassan G, Rosanova M, et al. Dissociations between spontaneous electroencephalographic features and the perturbational complexity index in the minimally conscious state. European Journal of Neuroscience. Published online March 1, 2024. doi:10.1111/ejn.16299
Casarotto S, Comanducci A, Rosanova M, et al. Stratification of unresponsive patients by an independently validated index of brain complexity. Ann Neurol. 2016;80(5):718-729. doi:10.1002/ana.24779
Sheybani L, Vivekananda U, Rodionov R, et al. Wake slow waves in focal human epilepsy impact network activity and cognition. Nat Commun. 2023;14(1). doi:10.1038/S41467-023-42971-3
Panet-Raymond D, Gotman J. Asymmetry in delta activity in patients with focal epilepsy. Electroencephalogr Clin Neurophysiol. 1990;75(6):474-481. doi:10.1016/0013-4694(90)90134-6
Vanrumste B, Jones RD, Bones PJ, Carroll GJ. Slow-wave activity arising from the same area as epileptiform activity in the EEG of paediatric patients with focal epilepsy. Clin Neurophysiol. 2005;116(1):9-17. doi:10.1016/J.CLINPH.2004.07.032
Blümcke I, Thom M, Aronica E, et al. The clinicopathologic spectrum of focal cortical dysplasias: a consensus classification proposed by an ad hoc Task Force of the ILAE Diagnostic Methods Commission. Epilepsia. 2011;52(1):158-174. doi:10.1111/J.1528-1167.2010.02777.X
Rocchi F, Canella C, Noei S, et al. Increased fMRI connectivity upon chemogenetic inhibition of the mouse prefrontal cortex. Nat Commun. 2022;13(1). doi:10.1038/s41467-022-28591-3
Baldassarre A, Ramsey LE, Siegel JS, Shulman GL, Corbetta M. Brain connectivity and neurological disorders after stroke. Curr Opin Neurol. 2016;29(6):706-713. doi:10.1097/WCO.0000000000000396
Avvenuti G, Handjaras G, Betta M, et al. Integrity of Corpus Callosum Is Essential for theCross-Hemispheric Propagation of Sleep Slow Waves:A High-Density EEG Study in Split-Brain Patients. J Neurosci. 2020;40(29):5589-5603. doi:10.1523/JNEUROSCI.2571-19.2020
Gao R. Interpreting the electrophysiological power spectrum. J Neurophysiol. 2016;115(2):628-630. doi:10.1152/JN.00722.2015
Lendner JD, Helfrich RF, Mander BA, et al. An electrophysiological marker of arousal level in humans. Elife. 2020;9:1-29. doi:10.7554/eLife.55092

Yes there is potential Competing Interest. Marcello Massimini is co-founder and shareholder of Intrinsic Powers, a spin-off of the University of Milan; Simone Sarasso is advisor of the same company. All other authors do not have any conflict of interest related to the present submission.

SupplementaryMaterialforNatComm.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Hemispherotomy: cortical islands of deep sleep in awake humans

Status:

Version 1

Abstract

Figures

Introduction

Materials and methods

Patients receiving disconnective surgery

Data acquisition and pre-processing

Estimation of slow delta power and spectral exponent

Statistical comparisons: spatial inter-hemispheric differences, and temporal pre-post differences

Spectral exponent: comparison to reference values during wake and sleep

Results

Patients and EEG recordings

Surgical disconnection and EEG slowing

Statistical comparisons across hemispheres and pre- post- surgery

Comparison of spectral exponent to a reference sample during wake and sleep

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1