Yuanfeng Zhou

The World Health Organization (WHO) defines brain health as the maintenance of optimal brain integrity, good mental state, and cognitive function without significant neuropsychiatric disease. Early childhood is a critical period for brain development, which were influenced by early experiences, nutrition, and environmental factors. Disruptions during this sensitive period can lead to lifelong challenges in cognition, emotion, and social function, highlighting early brain health as both a healthcare and societal priority. Despite growing global attention to brain health, resource distribution remains highly uneven. In particular, 43% of children in low‐ and middle‐income countries are at risk of not reaching their developmental potential. Only a few countries have implemented comprehensive, family‐friendly policies. To achieve the goal of a 30% reduction in neurodevelopmental disorders by 2030, a multidimensional strategy is essential, including providing balanced nutrition, promoting physical activity, enhancing mental health support, and creating safe and supportive society. Moreover, there is a need to accelerate the development of healthcare systems and promote the application of cutting‐edge technologies—such as molecular classification, multi‐omics research, and AI‐based prediction models. In response to these challenges, China is actively advancing its brain health agenda by establishing a National Center for Child Brain Health, launching large‐scale professional training programs, and promoting coordination between primary and comprehensive brain health centers. This review summarizes the major challenges in pediatric brain health, recent progress in related research, and key global and national strategies aimed at promoting optimal neurological development in children.

Purpose To evaluate the clinical efficacy and long-term functional prognosis of vagus nerve stimulation (VNS) combined with ketogenic diet (KD) in the treatment of children with febrile infection-related epilepsy syndrome (FIRES). Methods A retrospective analysis was performed on inpatients with a diagnosis of FIRES. They were treated with either VNS + KD (with an interval < 5 days) or KD alone, in addition to standard therapy. Results Eighteen children (nine females, nine males) with a median onset age of 8.1 years were studied. Before KD/VNS treatment, the average number of seizures (AS) per hour was 3.0 and the percentage of total duration of electrographic seizures (ESD) per hour was 65.5%. Median duration of anesthesia, duration of mechanical ventilation, and duration in ICU and hospital were 58.5, 39.5, 65.5, and 74.5 days, respectively. Four children received VNS + KD, and 14 received KD alone. The median time to intervention was 16.5 days after onset, and the median follow-up duration was 7.1 years. The VNS + KD group showed a significant reduction in average seizure frequency per hour (P = 0.002). However, no significant differences were observed in ESD or duration of status epilepticus (SE) (P > 0.05). And there were no significant differences between the groups in terms of duration of anesthesia, duration of mechanical ventilation, and duration in ICU and hospital (P > 0.05). At the last follow-up, 55.6% had moderate to severe disabilities, and 16.7% had died. The VNS + KD group exhibited significantly better functional outcomes as measured by the modified Rankin scale (P = 0.002). Younger age of onset was associated with increased odds of a poor outcome (P = 0.045). Conclusion FIRES patients experienced severe seizure burdens in the acute phase and had poor long-term prognoses. Combining VNS and KD treatments in the acute phase may represent a viable therapeutic strategy for reducing seizures and improving long-term function for pediatric FIRES.

Deep brain stimulation (DBS) has emerged as an important therapeutic intervention, effectively addressing a spectrum of drug-resistant neurological and psychiatric disorders. Although its efficacy has been validated in adult populations, the current literature reveals a significant gap concerning its application in pediatric patients. Specifically, pediatric populations afflicted with severe conditions such as dystonia, drug-resistant epilepsy, Tourette syndrome, and some other neuropsychiatric conditions demonstrate an urgent need for alternative therapeutic options. This review systematically examined the existing literature on the application of DBS in pediatric neurological disorders, focusing on the aforementioned conditions. Preliminary findings indicate that while DBS shows potential for a specific subset of pediatric patients, the current data is limited and lacks statistical power. Reported cases exhibit varying degrees of therapeutic success. Although adverse effects associated with DBS in pediatric populations are rare, further investigation is essential to define safety profiles accurately. Future research should focus on conducting large-scale, randomized controlled trials to validate outcomes and determine optimal patient selection criteria, thereby broadening its clinical application within the pediatric population.

Aims: Drug‐refractory epilepsy (DRE) refers to the failure of controlling seizures with adequate trials of two tolerated and appropriately chosen anti‐seizure medications (ASMs). For patients with DRE, surgical intervention becomes the most effective and viable treatment, but its success rate is unsatisfactory at only approximately 50%. Predicting surgical outcomes in advance can provide additional guidance to clinicians. Despite the high accuracy of invasive methods, they inevitably carry the risk of post‐operative infection and complications. Herein, to noninvasively predict surgical outcomes of DRE, we propose the “source causal connectivity” framework.Methods: In this framework, sLORETA, an EEG source imaging technique, was first used to inversely reconstruct intracranial neuronal electrical activity. Then, full convergent cross mapping (FCCM), a robust causal measure was introduced to calculate the causal connectivity between remodeled neuronal signals within epileptogenic zones (EZs). After that, statistical tests were performed to find out if there was a significant difference between the successful and failed surgical groups. Finally, a model for surgical outcome prediction was developed by combining causal network features with machine learning.Results: A total of 39 seizures with 205 ictal EEG segments were included in this prospective study. Experimental results exhibit that source causal connectivity in α‐frequency band (8~13 Hz) gains the most significant differences between the surgical success and failure groups, with a p‐value of 5.00e‐05 and Cohen's d effect size of 0.68. All machine learning models can achieve an average accuracy of higher than 85%. Among them, the SVM classifier with Gaussian kernel function and Bayesian optimization demonstrates the highest accuracy of 90.73%, with a PPV of 87.91%, an NPV of 92.98%, a sensitivity of 90.91%, a specificity of 90.60%, and an F1‐score of 89.39%.Conclusion: Our results demonstrate that the source causal network of EZ is a reliable biomarker for predicting DRE surgical outcomes. The findings promote noninvasive precision medicine for DRE.

Interictal epileptiform spikes (spikes) and epileptogenic focus are strongly correlated. However, partial spikes are insensitive to epileptogenic focus, which restricts epilepsy neurosurgery. Therefore, identifying spike subtypes that are strongly associated with epileptogenic focus (traceable spikes) could facilitate their use as reliable signal sources for accurately tracing epileptogenic focus. However, the sparse firing phenomenon in the transmission of intracranial neuronal discharges leads to differences within spikes that cannot be observed visually. Therefore, neuro-electro-physiologists are unable to identify traceable spikes that could accurately locate epileptogenic focus. Herein, we propose a novel sparse spike feature learning method to recognize traceable spikes and extract discrimination information related to epileptogenic focus. First, a multilevel eigensystem feature representation was determined based on a multilevel feature representation module to express the intrinsic properties of a spike. Second, the sparse feature learning module expressed the sparse spike multi-domain context feature representation to extract sparse spike feature representations. Among them, a sparse spike encoding strategy was implemented to effectively simulate the sparse firing phenomenon for the accurate encoding of the activity of intracranial neurosources. The sensitivity of the proposed method was 97.1%, demonstrating its effectiveness and significant efficiency relative to other state-of-the-art methods.

Objective: Patients with drug-resistant epilepsy (DRE) are commonly treated using neurosurgery, while its success rate is limited with approximately 50%. Predicting surgical outcomes is currently a prominent topic. The DRE is recognized as a network disorder involving a seizure triggering mechanism within epileptogenic zone (EZ); however, a systematic exploration of the EZ causal network remains lacking. Methods: This paper will advance DRE study by: (1) developing a novel causal coupling algorithm, “full convergent cross mapping (FCCM)” to improve the quantization performance; (2) characterizing the DRE's multi-frequency epileptogenic network by FCCM calculation of ictal iEEG; (3) predicting surgical outcomes using network features and machine learning. Numerical validations demonstrate the FCCM's superior quantization in terms of nonlinearity, accuracy, and stability. A multicenter cohort containing 22 DRE patients with 81 seizures is included. Result: Based on the Mann-Whitney-U-test, coupling strength of the epileptogenic network in successful surgeries is significantly higher than that of the failed group, with the most significant difference observed in α -iEEG network ( p = 1.52e - 07 ) Other clinical covariates are also considered and all th α -iEEG networks demonstrate consistent differences comparing successful and failed groups, with p = 0.014 and 9.23e - 06 for lesional and non-lesional DRE, p = 2.32e - 05, 0.0074 and 0.0030 for three clinical centers CHFU, JHU and NIH. Using FCCM features and 10-fold cross validation, the SVM achieves the highest accuracy of 87.65% in predicting surgical outcomes. Conclusion: The epileptogenic causal network is a reliable biomarker for estimating DRE's surgical outcomes. Significance: The proposed approach is promising to facilitate DRE precision medicine.

Neurodevelopmental disorders (NDDs) are a group of high-incidence rare diseases with genetic heterogeneity. PPP2R1A, the regulatory subunit of protein phosphatase 2A, is a recently discovered gene associated with NDDs. Whole/clinical exome sequencing was performed in five patients with a family with NDDs. In vitro experiments were performed to evaluate the mutants’ expression and interactions with the complex. The genotype–phenotype correlations of reported cases as well as our patients with PPP2R1A variants were reviewed. We reported five unrelated individuals with PPP2R1A variants, including two novel missense variants and one frameshift variant. The protein expression of the Arg498Leu variant was less than that of the wild-type protein, the frameshift variant Asn282Argfs*14 was not decreased but truncated, and these two variants impaired the interactions with endogenous PPP25RD and PPP2CA. Furthermore, we found that pathogenic variants clustered in HEAT repeats V, VI and VII, and patients with the Met180Val/Thr variants had macrocephaly, severe ID and hypotonia, but no epilepsy, whereas those with Arg258 amino acid changes had microcephaly, while a few had epilepsy or feeding problems. In this study, we reported five NDD patients with PPP2R1A gene variants and expanded PPP2R1A pathogenic variant spectrum. The genotype and phenotype association findings provide reminders regarding the prognostication and evidence for genetic counseling.

Background: Focal cortical dysplasia type 2 (FCD2) is a malformation of cortical development that constitutes a common cause of pediatric focal epilepsy. Germline or somatic variants in the mammalian target of rapamycin (mTOR) signaling pathway genes are the pathogenesis of FCD2.Objective: In this study, whole‐exome deep sequencing was performed on dysplastic cortex from focal epilepsy in children to explore genetic characteristics in FCD2.Methods: Resected core lesions of FCD2 were confirmed by pathology, and peripheral blood was collected from 11 patients. Deep whole‐exome sequencing (>500X) was performed on derived genomic DNA, germline, or somatic variants in brain‐specific genes were analyzed and identified.Results: In 11 patients, a heterozygous likely pathogenic germline variant of DEPDC5 was identified in one case, while somatic variants were found in four brain samples. The frequencies of the somatic variant allele were 2.52%–5.12%. Somatic variants in AKT3, TSC2, and MTOR (mTOR signaling pathway genes) were found in three samples. Besides, one somatic variant was detected in MED12 which has not been reported to associate with FCD2.Conclusion: Our study expanded the variant spectrum in the mTOR‐GATOR pathway, and also detected a somatic variant in MED12 which was potentially associated with FCD 2.

Objective . Mixing/dissociation of sleep stages in narcolepsy adds to the difficulty in automatic sleep staging. Moreover, automatic analytical studies for narcolepsy and multiple sleep latency test (MSLT) have only done automatic sleep staging without leveraging the sleep stage profile for further patient identification. This study aims to establish an automatic narcolepsy detection method for MSLT. Approach. We construct a two-phase model on MSLT recordings, where ambiguous sleep staging and sleep transition dynamics make joint efforts to address this issue. In phase 1, we extract representative features from electroencephalogram (EEG) and electrooculogram (EOG) signals. Then, the features are input to an EasyEnsemble classifier for automatic sleep staging. In phase 2, we investigate sleep transition dynamics, including sleep stage transitions and sleep stages, and output likelihood of narcolepsy by virtue of principal component analysis (PCA) and a logistic regression classifier. To demonstrate the proposed framework in clinical application, we conduct experiments on 24 participants from the Children’s Hospital of Fudan University, considering ten patients with narcolepsy and fourteen patients with MSLT negative. Main results. Applying the two-phase leave-one-subject-out testing scheme, the model reaches an accuracy, sensitivity, and specificity of 87.5%, 80.0%, and 92.9% for narcolepsy detection. Influenced by disease pathology, accuracy of automatic sleep staging in narcolepsy appears to decrease compared to that in the non-narcoleptic population. Significance. This method can automatically and efficiently distinguish patients with narcolepsy based on MSLT. It probes into the amalgamation of automatic sleep staging and sleep transition dynamics for narcolepsy detection, which would assist clinic and neuroelectrophysiology specialists in visual interpretation and diagnosis.