An Integrated Neurophysiological Approach Toward the Early Detection of Psychiatric Disorders

Schizophrenia (SCZ) is among the most debilitating syndromes within the neuropsychiatric domain, with patients suffering from this condition displaying a rich nosographic characterization. Besides, anhedonia, social withdrawal, and a progressive cognitive decline, positive manifestations such as delusions and hallucinations are perhaps the hallmark symptomatological features underlying this disorder. Notably, prodromal stages that entail subclinical cognitive and/or psychotic symptoms have often been shown to anticipate SCZ onset, corroborating a dimensional approach toward the understanding of such psychiatric disorder. Indeed, it has been proposed that schizotypal traits exist within the general population and that schizotypy and SCZ may correspond to quantitative poles within the same continuum, with the former (i.e., individuals characterized with high schizotypal traits, HST) including subjects displaying many SCZ-related subthreshold clinical signs, and thus, more likely to develop the latter. However, no reliable biomarker to early detect such psychosis-prone individuals has been so far, leaving clinicians without pre-emptive policies toward risk mitigation.

It should be nonetheless highlighted that both demeaned microstructural and white matter integrity have been reported in schizotypy and SCZ alike, the magnitude of which predicting the severity of positive symptoms at the individual level. Given the pivotal role these neuroanatomical features play in relaying electrical signals across the brain, it is not surprising that SCZ patients, as well as HST individuals, oftentimes exhibit aberrant oscillatory dynamics within the low-frequency ranges, such as in the alpha band (7-13 Hz). These brain oscillations are the dominant rhythm in both the resting and active brain. As such, they have been found to subserve many functions closely tied to sensory processing. Among multifold operational features these rhythms appear to be endowed with, their occipital cycling speed (i.e., individual alpha frequency peak, also known as IAF) has been recently associated with the efficiency by which sensory information is integrated or segregated, according to the task at hand . In agreement with the discrete sampling hypothesis of perception, the faster (or slower) these rhythms oscillate, the higher (or lower) the observers' visual sampling rate and their perceptual or cognitive accuracy.

Of note, this electrocortical index bears trait-like features that have been shown to parametrize perceptual performance, especially in multisensory contexts.

For instance, the propensity to report an illusory percept in audiovisual tasks such as the Sound-Induced Flash Illusion (SIFI) was found to negatively correlate with the IAF as recorded from occipital loci: the slower the IAF, the greater the number of illusory percepts and vice-versa. These results align with pieces of research showing a critical contribution of alpha oscillatory speed when individuals are asked to perform cross-modal simultaneity judgments and visuo-tactile tasks as well. Intriguingly, evidence for a tie between the IAF and multisensory temporal binding windows (TBW) has been further reinvigorated by data showing slower IAFs in patients suffering from schizophrenia or displaying schizotypal traits. Namely, those clinical and subclinical populations performing poorly at the SIFI, and, more broadly, multisensory paradigms, due to enlarged TBWs.

As for the neurocomputational mechanics subsiding this intricate pattern of alterations, it remains a matter of debate whether they might be driven either by an over-exploitation of auditory-relayed priors, or poor bottom-up sensory processing. Indeed, although alterations in sensory integration/segregation are a core hallmark of SCZ, it is not entirely clear whether such behavioral and cognitive impairments might be due to defective sensory processing (i.e. poor sensory input due to reduced feedforward connectivity), or aberrant perceptual bias mechanisms (i.e. disproportionately high top-down signaling), and to which extent they may be present at prodromal stages of the disease. To this aim, we will therefore investigate the performance of SCZ patients, and individuals displaying either high (HST) or low (LST) schizotypal traits, in a series of tasks relying on multisensory interaction. We will do so by implementing signal detection theory (SDT) matched with logistic fitting, to clarify at a finer-grained level whether alterations in multisensory processes can be accounted for by deficits at the level of sensory input (d'), top-down control (c), or both.