The application of artificial intelligence (AI) to resting-state electroencephalography (EEG) data represents a burgeoning frontier in schizophrenia diagnostics. However, the inferential validity of findings is frequently compromised by methodological heterogeneity. This systematic review rigorously evaluates the diagnostic efficacy and methodological robustness of AI-generated EEG biomarkers by implementing the Prediction model Risk Of Bias Assessment Tool for Artificial Intelligence (PROBAST+AI).

Following PRISMA guidelines, a systematic search of PubMed, Scopus, and Web of Science identified 43 primary studies encompassing 3,420 patients with schizophrenia or first-episode psychosis and 3,164 healthy controls (total N = 6,584). Methodological quality and validation strategies were assessed using the 34 signaling questions of the PROBAST+AI framework.

The aggregate mean diagnostic accuracy across the literature was 93.38%, with deep learning architectures demonstrating superior metrics relative to classical machine learning (96.83% vs. 87.68%). Critically, external validation was restricted to 4.7% of the included studies. A pronounced performance discrepancy was observed between subject-wise (88.22%) and segment-wise (98.14%) validation protocols, indicating pervasive data leakage. Overall, 76.7% of the analyzed studies were characterized by a high risk of bias, predominantly stemming from suboptimal data partitioning and inadequate independent verification.

Contemporary diagnostic accuracies appear to be artifactually elevated due to structural methodological deficiencies and data leakage. Consequently, current evidence does not support the immediate clinical implementation of AI-EEG biomarkers in schizophrenia. To facilitate translational progress, we propose a standardized methodological framework optimizing validation strategies.

Inappropriate diagnostic testing using electroencephalography (EEG) not only burdens the healthcare system but is especially concerning in Canada, where EEG demand is rising alongside a nationwide shortage of technologists. Understanding factors that influence yield, as well as patterns in requisition quality and completeness, may help improve both clinical utility and resource allocation.

We performed a single-center retrospective cohort study analyzing EEGs performed at Kingston Health Sciences Centre from March to August 2024. Demographic and clinical data were collected from paper EEG requisitions and digital reports. Diagnostic yield was assessed across clinical variables, physician specialty and patient setting using descriptive statistics, logistic regressions and odds ratios.

EEGs detected abnormalities in ∼40% of 1077 EEGs, mostly non-epileptiform. Yield was higher among older adults (60+ years), inpatients in critical care, emergency patients, and those on anti-seizure medications or sedatives. Indications like “neuro-prognostication” and “concern for non-convulsive status epilepticus” had high yield (100%), whereas “syncope/presyncope” had poor yield (12.8%). Sleep-deprived EEGs did not significantly differ in yield compared to routine EEGs. Over half (56%) of the EEG requisition forms were incomplete. Requisition completeness varied by physician specialty: neurologists were found least likely to complete requisition forms compared to other specialties.

By identifying clinical factors, physician specialties and settings associated with low diagnostic yield, interventions can be tailored to improve requisition practices and guide efficient resource allocation.

Neurological injury is common in neonates undergoing cardiac surgery with cardiopulmonary bypass, yet perioperative neuroimaging and neuromonitoring practices vary across centres. We aimed to characterise current practices and provider perceptions across Pediatric Cardiac Intensive Care Society-affiliated centres.

We conducted an online, cross-sectional survey of Pediatric Cardiac Intensive Care Society members between August and October 2024. Part 1, completed by medical directors, assessed institutional characteristics and perioperative practices. Part 2, completed by multidisciplinary providers, evaluated perceptions of utility and sufficiency.

A total of 143 responses from 60 institutions were analysed, including 38 medical directors. Most centres had a dedicated cardiac ICU (87%) and were academically affiliated (76%). Routine neuroimaging was more common preoperatively than postoperatively (87% vs. 32%), with head ultrasound most frequently used. Near-infrared spectroscopy was the most common neuromonitoring modality, while routine electroencephalography was reported at 18% of centres preoperatively and 50% postoperatively. Physicians were more likely than advanced practice providers and nurses to report practices as sufficient. Neurocritical care teams were present at 42% of centres but were not associated with greater use of neuroimaging or monitoring.

Significant variability exists in perioperative practices and provider perceptions, highlighting the need for multidisciplinary alignment and standardised guidelines.

Lennox-Gastaut syndrome (LGS) is characterized by drug-resistant seizures, cognitive impairment, and significant comorbidities, requiring lifelong care. Despite therapeutic advances, real-world management and system-level barriers in Canada are not well defined. A national survey was conducted to examine Canadian practices in LGS diagnostic approaches, therapies, comorbidity management, counseling practices, and transition services, aiming to identify gaps and guide quality improvement initiatives.

Following a literature review, a 41-item web-based survey was developed and distributed to neurologists and allied health workers via the Canadian League Against Epilepsy and the Canadian Neurological Sciences Federation.

Among 111 respondents (77% physicians; 79% academic practice), diagnostic confidence was limited: 46% believed LGS is both over- and underdiagnosed, and only 5% felt most patients are accurately identified. Adult diagnosis was considered particularly challenging (46%). Genetic testing for unknown etiologies was routine for 67% and, however, was more common among pediatric than adult neurologists (81% vs 41%, p < 0.001). Valproic acid (98%), clobazam (82%), and lamotrigine (75%) were the most common first-line anti-seizure medications (ASMs), yet 67% rated available treatments as inadequate. Reported treatment barriers included limited ASM efficacy (68%), adverse effects (38%), and comorbidity management challenges (37%). Corpus callosotomy was perceived as underutilized (71% overall), suggesting access disparities. Time and resource limitations hindered comorbidity screening. Only 23% reported access to formal transition programs.

Significant gaps remain in the diagnosis and management of LGS in Canada, particularly in treatment, comorbidity screening, and transition care. National guidelines, standardized pathways, and improved access to multidisciplinary services are needed to enhance outcomes.

Major depressive disorder (MDD) and bipolar disorder (BD) exhibit overlapping depressive presentations, making early differentiation difficult and misdiagnosis clinically consequential. Entropy-based resting-state electroencephalography (EEG) markers were evaluated for differentiating drug-naïve patients with stable MDD, converted BD, and diagnosed BD.

Resting-state EEG was acquired from 92 drug-naïve patients: 50 with stable MDD, 16 who converted to bipolar II disorder over 3 years, and 26 with diagnosed BD. Band-scale entropy and cross-sample entropy (CSE) were analyzed. Group differences were tested by analysis of covariance adjusted for depressive severity; classification was evaluated by nested cross-validation with permutation testing.

Relative to stable MDD, diagnosed BD showed higher theta- and alpha-band entropy in frontal and central regions and lower values for selected intra- and inter-regional CSE measures. Two CSE measures correlated positively with non-planning impulsiveness in diagnosed BD. Converted BD showed intermediate values without significant pairwise differences. Prospective discrimination of stable MDD versus converted BD was modest (balanced accuracy, 65.4%; area under the receiver operating characteristic curve [ROC-AUC], 0.661). Classification was strongest for stable MDD versus diagnosed BD, yielding 78.9% pooled accuracy, 76.6% balanced accuracy, and an ROC-AUC of 0.832 when entropy features were combined with depressive severity; combined entropy outperformed either single-family configuration.

Entropy-derived EEG measures differentiated stable MDD from diagnosed BD and suggested a weaker, intermediate profile in converted BD. They may serve as candidate adjunctive biomarkers for mood disorder differentiation, although larger prospective cohorts are required to clarify their value in predicting conversion to bipolar II disorder.

Alterations in reward responsiveness represent a key mechanism implicated in youth depression risk. However, not all youth with these alterations develop depression, suggesting the presence of factors that may moderate risk patterns. As socioeconomic disadvantage is also related to youth depression risk, particularly for youth exhibiting altered reward function, this study examined whether indices of family- and neighborhood-level disadvantage interacted with electrocortical reward responsivity to predict depression symptom trajectories across childhood and adolescence.

Participants included 76 youth (ages 9–16 years) at low and high risk for depression based on maternal history of depression. At baseline, youth completed a monetary reward-guessing task while electroencephalography was recorded to measure the reward positivity (RewP), an event-related potential indexing reward responsiveness. Family and neighborhood disadvantage were assessed using the income-to-needs (ITN) ratio and Area Deprivation Index (ADI), respectively. Self-reported and clinician-rated depression symptoms were assessed across a multiwave, 18-month follow-up.

RewP interacted with family- and neighborhood-level disadvantage to predict self-reported depression symptom trajectories. Specifically, blunted RewP predicted self-reported depression symptom increases for youth with a lower ITN ratio and higher ADI score. A blunted RewP also predicted clinician-rated depression symptom increases for youth living in neighborhoods with higher ADI scores.

Findings suggest that reduced reward responsiveness is a mechanism implicated in future depression risk among youth, specifically in the context of family- and neighborhood-level socioeconomic disadvantage. Interventions that enhance reward response among youth exposed to higher levels of socioeconomic disadvantage may be particularly effective in preventing depression emergence.

Based on facial expression experiments, childhood adversity may be associated with threat-related information processing bias. Yet, it is unclear whether this generalizes to other threat-related stimuli, such as social and non-social visual scenes.

We combined fast periodic visual stimulation with frequency-tagging electroencephalography (EEG) and eye-tracking to assess automatic and implicit neural discrimination, neural salience and preferential looking towards negative versus neutral social and non-social visual scenes in young adults aged 16–24 years (51 with childhood adversity and psychiatric symptoms and 43 controls).

Controls showed enhanced negative-neutral neural discrimination within a social versus non-social context. However, this facilitating effect of social content was absent in those with adversity, suggesting a selective alteration in social threat processing. Moreover, individual differences in adversity severity, and more specifically threat experiences (but not neglect experiences), were associated with decreased neural discrimination of negative versus neutral social scenes, corresponding to similar findings in facial expression processing, indicating the robustness of adversity-related deficits in threat-safety discrimination across social visual stimuli.

The adversity-related decreased threat-safety discrimination might impact individuals’ perception of social cues in daily life and relate to poor social functioning and future development of psychopathology.

Recent neuroimaging studies have demonstrated that the heterogeneous antidepressant responsiveness in patients with major depressive disorder (MDD) is associated with diverse resting-state functional brain network (rsFBN) topology; however, only limited studies have explored the rsFBN using electroencephalography (EEG). In this study, we aimed to identify EEG-derived rsFBN-based biomarkers to predict pharmacotherapeutic responsiveness.

The resting-state EEG signals were acquired for demography-matched three groups: 98 patients with treatment-refractory MDD (trMDD), 269 those with good-responding MDD (grMDD), and 131 healthy controls (HCs). The source-level rsFBN was constructed using 31 sources as nodes and beta-band power envelope correlation (PEC) as edges. The degree centrality (DC) and clustering coefficients (CCs) were calculated for various sparsity levels. Network-based statistic and one-way analysis of variance models were employed for comparing PECs and network indices, respectively. The multiple comparisons were controlled by the false discovery rate.

Patients with trMDD were characterized by the altered dorsal attention network and salience network. Specifically, they exhibited hypoconnection between eye fields and right parietal regions (p = 0.0088), decreased DC in the right supramarginal gyrus (q = 0.0057), and decreased CC in the reward circuit (qs < 0.05). On the other hand, both MDD groups shared increased DC but decreased CC in the posterior cingulate cortex.

We confirmed that network topology was more severely deteriorated in patients with trMDD, particularly for the attention-regulatory networks. Our findings suggested that the altered rsFBN topologies could serve as potential pathologically interpretable biomarkers for predicting antidepressant responsiveness.

Nearly 25% of people with intellectual disability (PwID) have epilepsy compared to 1% of the UK general population. PwID are commonly excluded from research, eventually affecting their care. Understanding seizures in PwID is particularly challenging because of reliance on subjective external observation and poor objective validation. Remote electroencephalography (EEG) monitoring could capture objective data, but particular challenges and implementation strategies for this population need to be understood.

This co-production aimed to explore the accessibility and potential impact of a remote, long-term EEG tool (UnEEG 24/7 SubQ) for PwID and epilepsy.

We conducted six, 2-hour long workshops; three with people with mild intellectual disability and three with families/carers of people with moderate–profound intellectual disability. Brief presentations, easy read information and model demonstrations were used to explain the problem and device. A semi-structured guide developed by a communication specialist and art-based techniques facilitated discussion with PwID. For family/carers, active listening was employed. All conversations were recorded and transcribed. Artificial intelligence-based coding and thematic analysis (ATLAS.ti and ChatGPT) were synthesised with manual theming to generate insights.

Co-production included four PwID, five family members and seven care professionals. Three main themes were identified: (1) perceived benefits for improving seizure understanding, informing care and reducing family and carer responsibility to accurately identify seizures; (2) the device was feasible for some PwID but not all; and (3) appropriate person-centred communication is essential for all stakeholders to reduce concerns.

The workshops identified key benefits and implementing barriers to SubQ in PwID.

The Hierarchical Taxonomy of Psychopathology (HiTOP) offers a promising framework to identify the neurobiological mechanisms of psychopathology. Many forms of psychopathology are characterized by dysfunctional emotional reactivity. The late positive potential (LPP) is an event-related potential component that provides an index of neurobiological emotional reactivity. Several categorical disorders have demonstrated a similar association with the emotion-modulated LPP. It is possible that higher-order dimensional representations of psychopathology might explain the comparable results. The present study examined the association between HiTOP-consistent pathological personality dimensions across multiple levels of the hierarchy and neurobiological emotional reactivity.

The sample included 215 18–35-year-old adults (86% female) who were oversampled for psychopathology. Participants completed the emotional interrupt task while electroencephalography was recorded to examine the LPP. Participants also completed the Comprehensive Assessment of Traits relevant to Personality Disorders to assess pathological personality.

At the spectra level, higher negative emotionality was associated with a larger emotion-modulated LPP, while higher detachment was associated with a smaller emotion-modulated LPP. There were no associations between higher-order psychopathology levels and the emotion-modulated LPP. Compared to categorical diagnoses, spectra-level personality pathology dimensions significantly improved the prediction of the emotion-modulated LPP.

The present study indicates that HiTOP spectra levels of negative emotionality and detachment demonstrate unique associations with neurobiological emotional reactivity. The study highlights the utility of examining dimensional and hierarchical, rather than categorical, representations of psychopathology in the attempt to identify the neurobiological origins of psychopathology.