Exploring Temporal Stability and Consciousness: A Novel Approach to Identifying Cognitive Motor Dissociation

Consciousness, the enigmatic cornerstone of our intricate mental existence, has long captivated the realms of neuroscience, with its underlying neural mechanisms an enduring enigma (Northoff and Lamme, 2020; Seth and Bayne, 2022).

Among the multifaceted avenues to comprehend this profound phenomenon, the exploration of disorders of consciousness (DOC) stands as a prominent approach.

In the realm of clinical neuroscience, DOC, often stemming from severe brain injuries, is partitioned into distinct categories: the unresponsive wakefulness syndrome (UWS), colloquially known as the vegetative state, and the minimally conscious state (MCS) (Giacino et al., 2014; Sanz et al., 2021).

UWS encapsulates individuals devoid of awareness, exhibiting no cognitive connection to their surroundings or selves (Laureys et al., 2010).

In contrast, MCS encompasses individuals displaying intermittent yet reproducible signs of consciousness (Giacino et al., 2002). The evolution of our comprehension of disorders of consciousness has birthed a paradigm-shifting revelation—a subset of DOC patients (approximately 20%) retaining covert awareness (Claassen et al., 2019; Monti et al., 2010; Owen et al., 2006).

This intriguing cohort, coined as having cognitive motor dissociation (CMD) by Schiff (2015), grapples with the paradoxical predicament of losing the ability to execute purposeful movements while preserving a rich tapestry of conscious experiences (Schiff, 2015).

Notably, CMD patients exhibit a distinct trajectory of improvement when contrasted with traditional DOC patients (Edlow et al., 2021; Jöhr et al., 2020), magnifying the significance of an accurate CMD diagnosis, a determinant with direct implications on therapeutic strategies and rehabilitation outcomes (Thibaut et al., 2019).

To this end, two prevailing methodologies emerge to diagnose CMD patients: the task-based neuroimaging technique encompassing EEG and fMRI (Claassen et al., 2019; Monti et al., 2010), and the innovative brain-computer interface (BCI) method (Gibson et al., 2016; Pan et al., 2020).

Both methodologies pivot on patients’ capacity to comprehend and execute oral commands, precipitating a dilemma in cases where cognitive functions, such as attention and memory, are impaired (Edlow et al., 2017; Sanz et al., 2021). Herein lies the potential for misdiagnosis – a CMD patient’s covert awareness might lead to erroneous conclusions of unconsciousness due to the inability to execute specific tasks with precision.

This predicament illuminates the need for alternative methodologies, prompting the spotlight on resting-state fMRI (rs-fMRI)—a compelling non-invasive technique that sidesteps task-related limitations while capturing brain activity (Battaglia et al., 2020; Jie et al., 2018; Zhang et al., 2018). Yet, empirical evidence substantiating the diagnostic prowess of rs-fMRI remains elusive.

Traditional investigations into DOC via neuroimaging predominantly adopted the static rs-fMRI functional connectivity (FC) methodology, unveiling disrupted or reorganized functional networks within the DOC population (Qin et al., 2021; Sinitsyn et al., 2018; Vanhaudenhuyse et al., 2010; Wu et al., 2015).

However, the notion of stationary brain function has been challenged by a growing body of evidence that underscores the brain’s dynamic nature (Preti et al., 2017; Thompson and Fransson, 2018). A paradigm shift emerges with the introduction of dynamic functional connectivity (dFC) analyses, exposing irregularities in brain dynamics underlying DOC (Demertzi et al., 2019; Huang et al., 2020; Luppi et al., 2019).

Against this backdrop, the current study diverges from the customary multiple discrete states premise in brain dynamics (Allen et al., 2014), favoring an exploration of dynamic functional organizations as a continuous process, accentuating the significance of temporal stability (Battaglia et al., 2020; Jie et al., 2018; Zhang et al., 2018).

The underpinning notion is that sustained awareness hinges on a stable functional architecture over time, marked by a consistent, widespread, and reproducible connectivity pattern (Dehaene et al., 2017; Li et al., 2020). This temporal stability may be jeopardized in various mental disorders, including Alzheimer’s disease, schizophrenia, and autism (Easson and McIntosh, 2019; Rolls et al., 2021; Zhang et al., 2018), all of which are characterized by cognitive function impairment—a bedrock of normal awareness (Berkovitch et al., 2017; Leekam, 2016; O’Shaughnessy et al., 2021).

The convergence of research underscores the paramount significance of temporal stability for consciousness. However, empirical evidence remains sparse regarding how temporal stability shifts within the context of impaired awareness in DOC. Furthermore, the extent to which CMD patients’ preserved awareness aligns with the temporal stability pattern of fully conscious individuals remains an intriguing puzzle.

This study embarks on a dual voyage: firstly, to scrutinize the altered temporal stability in DOC with impaired awareness, constructing a linear support vector machine (SVM) classifier to differentiate fully conscious states from impaired conscious states (DOC); secondly, to assess the SVM model’s accuracy in identifying CMD patients harboring covert awareness.

The study materializes through the amalgamation of two independent datasets. The inaugural dataset (n = 52) comprises 16 fully conscious participants with a history of brain injury and 36 DOC patients. This dataset serves as the crucible for training an SVM classification model tasked with distinguishing DOC patients from controls.

The model’s mettle undergoes rigorous testing within the second dataset (n = 56), which houses 21 healthy controls and 35 DOC patients. The second facet of the study necessitates a third dataset (n = 11), accommodating 4 CMD patients and 7 potential non-CMD individuals.

The critical distinction lies in these patients’ awareness levels, discerned in a prior EEG-based BCI study (Pan et al., 2020). Capitalizing on this insight, the SVM model is transposed onto the third dataset, an arena where its capacity to classify CMD patients as fully conscious controls and potential non-CMD subjects as DOC cases is scrutinized (Fig. 1).

The study’s findings unveil impaired global and regional temporal stability among DOC patients relative to controls. Crucially, the SVM model not only demonstrates proficiency in discriminating between DOC patients and controls within the initial and second datasets (with an accuracy of 90% and 84% respectively), but also excels in distinguishing between CMD patients and potential non-CMD individuals in the third dataset (accuracy = 91%).

This multi-pronged investigation offers invaluable insights into the intricate dynamics of consciousness, all the while spotlighting the potential of temporal stability as a diagnostic beacon within the realm of cognitive motor dissociation.

In summation, the study beckons us to traverse the intricate tapestry of cognitive motor dissociation and temporal stability within patients grappling with disorders of consciousness.

In its quest to redefine traditional paradigms and harness cutting-edge methodologies, the research not only elucidates the landscape of temporal stability within impaired consciousness but also introduces a pioneering SVM-based classifier that transcends conventional diagnostic horizons in identifying cognitive motor dissociation.

This journey holds the promise of refined diagnostic precision and transformative therapeutic interventions within the captivating realm of consciousness research.

reference link : https://www.sciencedirect.com/science/article/pii/S1053811923001969