Perinatal arterial ischemic stroke (PAIS) is a major cause of long-term neurological impairments. Understanding its etiologic and prognostic factors is essential for improving outcomes.
Methods
A retrospective single-center study was conducted on children diagnosed with PAIS between 2008 and 2024. Clinical, imaging, and EEG data were analyzed to identify predictors of epilepsy and poor outcomes based on modified Rankin Scale (mRS) scores.
Results
A total of 44 patients (9 neonatal arterial ischemic stroke [NAIS], 35 Presumed perinatal arterial ischemic stroke [PPAIS]) with a mean follow-up of 6.49 years were included. Seizure was the first symptom in 45.5%, and focal motor deficit in 54.5%. The most affected artery was the middle cerebral artery (92%), particularly the M1 branch (36.4%). Cortical involvement was noted in 59.1%. Fourteen patients developed epilepsy; 14.2% were drug-resistant. Focal EEG slowing (odds ratio [OR] = 8.484, p = 0.019) and cortical involvement (OR = 6.857, p = 0.023) significantly predicted epilepsy. Poor mRS outcomes (≥3) were seen in 44.2%, and were associated with epilepsy (OR = 5.556, p = 0.016), EEG slowing (OR = 5.353, p = 0.035), and cortical involvement (OR = 7.467, p = 0.008).
Conclusions
Early EEG and MRI findings are crucial in predicting long-term prognosis in PAIS. Further multicenter studies are needed to validate and enhance outcomes.
背景围产期动脉缺血性中风（PAIS）是导致长期神经功能损伤的主要原因。了解其病因和预后因素对改善预后至关重要。方法对2008 ~ 2024年诊断为PAIS的儿童进行回顾性单中心研究。对临床、影像学和脑电图数据进行分析，以确定基于改良兰金量表（mRS）评分的癫痫和不良预后的预测因素。结果共纳入44例患者，其中新生儿动脉缺血性卒中9例，围产期动脉缺血性卒中35例，平均随访6.49年。癫痫是45.5%的第一症状，54.5%为局灶性运动障碍。受影响最大的动脉为大脑中动脉（92%），尤其是M1分支（36.4%）。59.1%的患者有皮层受累。14例发生癫痫；14.2%耐药。局灶性脑电图减慢（优势比[OR] = 8.484, p = 0.019）和皮层受累（OR = 6.857, p = 0.023）与癫痫有显著相关性。44.2%的患者出现不良mRS结果（≥3），并与癫痫（OR = 5.556, p = 0.016）、脑电图减慢（OR = 5.353, p = 0.035）和皮层受累（OR = 7.467, p = 0.008）相关。结论早期脑电图和MRI表现对预测PAIS的远期预后至关重要。需要进一步的多中心研究来验证和增强结果。
{"title":"Evaluation of perinatal arterial ischemic stroke patients: Underlying etiologic factors and long-term prognosis","authors":"Çisem Duman Kayar , Rıdvan Avcı , Fulya Kürekçi , Ceyda Öney Yılmaz , Ceyda Bayraktar Eltutan , Vugar Abbasaliyev , Serap Karaman , Mehmet Barbüroğlu , Hülya Maraş Genç , Edibe Pembegül Yıldız","doi":"10.1016/j.braindev.2026.104501","DOIUrl":"10.1016/j.braindev.2026.104501","url":null,"abstract":"<div><h3>Background</h3><div>Perinatal arterial ischemic stroke (PAIS) is a major cause of long-term neurological impairments. Understanding its etiologic and prognostic factors is essential for improving outcomes.</div></div><div><h3>Methods</h3><div>A retrospective single-center study was conducted on children diagnosed with PAIS between 2008 and 2024. Clinical, imaging, and EEG data were analyzed to identify predictors of epilepsy and poor outcomes based on modified Rankin Scale (mRS) scores.</div></div><div><h3>Results</h3><div>A total of 44 patients (9 neonatal arterial ischemic stroke [NAIS], 35 Presumed perinatal arterial ischemic stroke [PPAIS]) with a mean follow-up of 6.49 years were included. Seizure was the first symptom in 45.5%, and focal motor deficit in 54.5%. The most affected artery was the middle cerebral artery (92%), particularly the M1 branch (36.4%). Cortical involvement was noted in 59.1%. Fourteen patients developed epilepsy; 14.2% were drug-resistant. Focal EEG slowing (odds ratio [OR] = 8.484, <em>p</em> = 0.019) and cortical involvement (OR = 6.857, <em>p</em> = 0.023) significantly predicted epilepsy. Poor mRS outcomes (≥3) were seen in 44.2%, and were associated with epilepsy (OR = 5.556, <em>p</em> = 0.016), EEG slowing (OR = 5.353, <em>p</em> = 0.035), and cortical involvement (OR = 7.467, <em>p</em> = 0.008).</div></div><div><h3>Conclusions</h3><div>Early EEG and MRI findings are crucial in predicting long-term prognosis in PAIS. Further multicenter studies are needed to validate and enhance outcomes.</div></div>","PeriodicalId":56137,"journal":{"name":"Brain & Development","volume":"48 1","pages":"Article 104501"},"PeriodicalIF":1.3,"publicationDate":"2026-01-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145977867","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
MRI-based classification of hypoxic–ischemic encephalopathy (HIE), particularly using diffusion-weighted imaging (DWI), remains susceptible to subjective bias in determining the presence or absence of diffusion restriction. Quantitative measures such as apparent diffusion coefficient (ADC) values may provide a more objective assessment of injury severity.
Purpose
To evaluate the depth and distribution of brain injury in neonatal HIE by quantitatively assessing regional ADC values across MRI pattern classifications, independent of subjective visual interpretation of DWI signal changes.
Materials and methods
Seventy-three consecutively enrolled full-term infants with HIE underwent brain MRI, and ADC values were measured at five predefined brain regions. Correlations between regional ADC values and MRI pattern groups were analyzed.
Results
Five MRI patterns were identified: normal, white matter injury (WMI), watershed (WS), basal ganglia/thalamus injury (BGT), and near total brain injury (nTBI). ADC distribution analysis demonstrated two principal findings: (1) no significant differences among the normal, WMI, and WS pattern groups; and (2) significantly lower ADC values in the nTBI group compared with the BGT group at all brain regions during the first week of life, and at the centrum semiovale during the second week.
Conclusion
Quantitative analysis revealed distinct ADC distribution differences between the BGT and nTBI patterns during the first and second weeks of life, indicating substantially more severe and widespread brain injury in the nTBI pattern.
{"title":"Quantitative interpretation using ADC values for subjective MRI classification of neonatal hypoxic-ischemic encephalopathy","authors":"Katsumi Hayakawa , Koichi Tanda , Masakazu Nishimoto , Akira Nishimura , Daisuke Kinoshita , Yuko Sano","doi":"10.1016/j.braindev.2025.104499","DOIUrl":"10.1016/j.braindev.2025.104499","url":null,"abstract":"<div><h3>Background</h3><div>MRI-based classification of hypoxic–ischemic encephalopathy (HIE), particularly using diffusion-weighted imaging (DWI), remains susceptible to subjective bias in determining the presence or absence of diffusion restriction. Quantitative measures such as apparent diffusion coefficient (ADC) values may provide a more objective assessment of injury severity.</div></div><div><h3>Purpose</h3><div>To evaluate the depth and distribution of brain injury in neonatal HIE by quantitatively assessing regional ADC values across MRI pattern classifications, independent of subjective visual interpretation of DWI signal changes.</div></div><div><h3>Materials and methods</h3><div>Seventy-three consecutively enrolled full-term infants with HIE underwent brain MRI, and ADC values were measured at five predefined brain regions. Correlations between regional ADC values and MRI pattern groups were analyzed.</div></div><div><h3>Results</h3><div>Five MRI patterns were identified: normal, white matter injury (WMI), watershed (WS), basal ganglia/thalamus injury (BGT), and near total brain injury (nTBI). ADC distribution analysis demonstrated two principal findings: (1) no significant differences among the normal, WMI, and WS pattern groups; and (2) significantly lower ADC values in the nTBI group compared with the BGT group at all brain regions during the first week of life, and at the centrum semiovale during the second week.</div></div><div><h3>Conclusion</h3><div>Quantitative analysis revealed distinct ADC distribution differences between the BGT and nTBI patterns during the first and second weeks of life, indicating substantially more severe and widespread brain injury in the nTBI pattern.</div></div>","PeriodicalId":56137,"journal":{"name":"Brain & Development","volume":"48 1","pages":"Article 104499"},"PeriodicalIF":1.3,"publicationDate":"2026-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145977866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Epilepsy is a genetically heterogeneous disorder with a high burden in the pediatric population. Advances in next-generation sequencing (NGS) have enhanced molecular diagnosis, enabling more accurate subclassification and targeted interventions.
Objective
This study aimed to evaluate the diagnostic utility of epilepsy gene panel testing in a large pediatric cohort and to characterize the clinical and genetic features of molecularly diagnosed cases.
Methods
A retrospective analysis was conducted on 516 pediatric epilepsy patients who underwent targeted gene panel testing between 2021 and 2025 at a tertiary medical center. Only pathogenic and likely pathogenic variants were considered diagnostic.
Results
A molecular diagnosis was established in 81 patients (15.7 %). The most frequently implicated genes were KCNQ2, SCN1A, CACNA1A, SLC2A1, and SCN2A, collectively accounting for 43.2 % of all diagnoses. Despite this concentration, pathogenic variants were distributed across 37 different genes, emphasizing the high genetic heterogeneity. Most diagnosed patients had seizure onset in infancy, particularly within the first year of life. Notably, two-thirds of neonates with seizures had pathogenic KCNQ2 variants. Additionally, 23 of the diagnostic variants (29.9 %) were novel, underscoring the evolving spectrum of epilepsy-associated mutations.
Conclusion
Epilepsy gene panel testing is a valuable diagnostic tool in pediatric clinical practice. The identification of pathogenic variants across a wide range of genes — including a high proportion of novel mutations — supports the integration of genetic testing into the routine evaluation of pediatric epilepsy for improved etiological clarification and long-term management.
{"title":"Targeted gene panel testing in pediatric epilepsy: Diagnostic outcomes and expanding genetic insights","authors":"Aslihan Sanri , Unal Akca , Mehmet Burak Mutlu , Ozlem Sezer , Emre Sanri , Taner Karakaya , Serkan Kurtgoz","doi":"10.1016/j.braindev.2025.104498","DOIUrl":"10.1016/j.braindev.2025.104498","url":null,"abstract":"<div><h3>Background</h3><div>Epilepsy is a genetically heterogeneous disorder with a high burden in the pediatric population. Advances in next-generation sequencing (NGS) have enhanced molecular diagnosis, enabling more accurate subclassification and targeted interventions.</div></div><div><h3>Objective</h3><div>This study aimed to evaluate the diagnostic utility of epilepsy gene panel testing in a large pediatric cohort and to characterize the clinical and genetic features of molecularly diagnosed cases.</div></div><div><h3>Methods</h3><div>A retrospective analysis was conducted on 516 pediatric epilepsy patients who underwent targeted gene panel testing between 2021 and 2025 at a tertiary medical center. Only pathogenic and likely pathogenic variants were considered diagnostic.</div></div><div><h3>Results</h3><div>A molecular diagnosis was established in 81 patients (15.7 %). The most frequently implicated genes were <em>KCNQ2</em>, <em>SCN1A</em>, <em>CACNA1A</em>, <em>SLC2A1</em>, and <em>SCN2A</em>, collectively accounting for 43.2 % of all diagnoses. Despite this concentration, pathogenic variants were distributed across 37 different genes, emphasizing the high genetic heterogeneity. Most diagnosed patients had seizure onset in infancy, particularly within the first year of life. Notably, two-thirds of neonates with seizures had pathogenic <em>KCNQ2</em> variants. Additionally, 23 of the diagnostic variants (29.9 %) were novel, underscoring the evolving spectrum of epilepsy-associated mutations.</div></div><div><h3>Conclusion</h3><div>Epilepsy gene panel testing is a valuable diagnostic tool in pediatric clinical practice. The identification of pathogenic variants across a wide range of genes — including a high proportion of novel mutations — supports the integration of genetic testing into the routine evaluation of pediatric epilepsy for improved etiological clarification and long-term management.</div></div>","PeriodicalId":56137,"journal":{"name":"Brain & Development","volume":"48 1","pages":"Article 104498"},"PeriodicalIF":1.3,"publicationDate":"2026-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925918","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Primary coenzyme Q10 (CoQ10) deficiency is a group of mitochondrial disorders caused by pathogenic variants of genes involved in CoQ10 biosynthesis. Although some patients respond to oral CoQ10 supplementation, the pathophysiology remains poorly understood. Ferroptosis, a form of iron-dependent cell death driven by lipid peroxidation, is suppressed by reduced CoQ10via ferroptosis suppressor protein 1 (FSP1). However, its involvement in primary CoQ10 deficiency has not yet been studied using patient-derived cells.
Cases and results
We reported six patients from three families and investigated ferroptosis susceptibility in fibroblasts from three representative patients: one with COQ2 variants and two with COQ4 variants. Fibroblasts with COQ2 variants showed increased vulnerability to ferroptosis inducers, plasma membrane lipid peroxidation. In contrast, fibroblasts with COQ4 variants exhibited only mild changes. Notably, susceptibility to ferroptosis remained unchanged after increasing intracellular CoQ10 levels. Despite this persistent ferroptosis sensitivity in vitro, the COQ2 patient exhibited significant clinical improvement following CoQ10 supplementation. These findings suggest that ferroptosis may contribute to cellular vulnerability in primary CoQ10 deficiency but may not be the primary driver of renal and neurological symptoms.
Conclusions
Our results highlight a complex interplay between CoQ10 biosynthesis, ferroptosis defense, and therapeutic response, warranting further investigation of subcellular CoQ10 distribution and ferroptosis-related mechanisms.
{"title":"Ferroptosis susceptibility in primary coenzyme Q10 deficiency: Cellular insights from patient fibroblasts and clinical course of six individuals","authors":"Chika Watanabe , Akihiko Miyauchi , Shiho Aoki , Miyuki Watanabe , Eriko F. Jimbo , Yuudai Miyama , Hirotsugu Kitayama , Yuichi Uno , Kenji Watanabe , Yuka Hattori , Yuka Yotsumoto , Takanori Onuki , Yohei Sugiyama , Keiko Ichimoto , Yukiko Yatsuka , Yasushi Okazaki , Toshiyuki Imasawa , Kei Murayama , Akira Ohtake , Takanori Yamagata , Hitoshi Osaka","doi":"10.1016/j.braindev.2025.104497","DOIUrl":"10.1016/j.braindev.2025.104497","url":null,"abstract":"<div><h3>Background</h3><div>Primary coenzyme Q<sub>10</sub> (CoQ<sub>10</sub>) deficiency is a group of mitochondrial disorders caused by pathogenic variants of genes involved in CoQ<sub>10</sub> biosynthesis. Although some patients respond to oral CoQ<sub>10</sub> supplementation, the pathophysiology remains poorly understood. Ferroptosis, a form of iron-dependent cell death driven by lipid peroxidation, is suppressed by reduced CoQ<sub>10</sub> <em>via</em> ferroptosis suppressor protein 1 (FSP1). However, its involvement in primary CoQ<sub>10</sub> deficiency has not yet been studied using patient-derived cells.</div></div><div><h3>Cases and results</h3><div>We reported six patients from three families and investigated ferroptosis susceptibility in fibroblasts from three representative patients: one with <em>COQ2</em> variants and two with <em>COQ4</em> variants. Fibroblasts with <em>COQ2</em> variants showed increased vulnerability to ferroptosis inducers, plasma membrane lipid peroxidation. In contrast, fibroblasts with <em>COQ4</em> variants exhibited only mild changes. Notably, susceptibility to ferroptosis remained unchanged after increasing intracellular CoQ<sub>10</sub> levels. Despite this persistent ferroptosis sensitivity <em>in vitro</em>, the <em>COQ2</em> patient exhibited significant clinical improvement following CoQ<sub>10</sub> supplementation. These findings suggest that ferroptosis may contribute to cellular vulnerability in primary CoQ<sub>10</sub> deficiency but may not be the primary driver of renal and neurological symptoms.</div></div><div><h3>Conclusions</h3><div>Our results highlight a complex interplay between CoQ<sub>10</sub> biosynthesis, ferroptosis defense, and therapeutic response, warranting further investigation of subcellular CoQ<sub>10</sub> distribution and ferroptosis-related mechanisms.</div></div>","PeriodicalId":56137,"journal":{"name":"Brain & Development","volume":"48 1","pages":"Article 104497"},"PeriodicalIF":1.3,"publicationDate":"2025-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145866425","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Reply to “Comment on ‘Survival motor neuron protein is the optimal biomarker for evaluating the risdiplam treatment’”","authors":"Tamaki Kato, Noriko Otsuki, Mamoru Yokomura, Kayoko Saito","doi":"10.1016/j.braindev.2025.104493","DOIUrl":"10.1016/j.braindev.2025.104493","url":null,"abstract":"","PeriodicalId":56137,"journal":{"name":"Brain & Development","volume":"48 1","pages":"Article 104493"},"PeriodicalIF":1.3,"publicationDate":"2025-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145822195","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-13DOI: 10.1016/j.braindev.2025.104494
Gül Yücel , Nur Yücel Ekici
{"title":"Diagnostic advances in the etiology of cytotoxic lesions of the corpus callosum (CLOCC): epilepsy relationship and ADC radiomics perspective","authors":"Gül Yücel , Nur Yücel Ekici","doi":"10.1016/j.braindev.2025.104494","DOIUrl":"10.1016/j.braindev.2025.104494","url":null,"abstract":"","PeriodicalId":56137,"journal":{"name":"Brain & Development","volume":"48 1","pages":"Article 104494"},"PeriodicalIF":1.3,"publicationDate":"2025-12-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738495","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-11DOI: 10.1016/j.braindev.2025.104491
Susanne R. De Rooij , Amber Boots
Background
There is a large body of literature on brain health, usually focusing on brain health in adulthood in relation to neurodegenerative diseases. Over the past ten years, brain health has also emerged as a concept in studies conducted in youth. But what is brain health in youth and how can we measure this?
Methods
We summarized the literature on brain health in youth and inventoried operationalizations of brain health and research themes. We searched Medline for studies reporting on brain health in youth (from neonates to late adolescent stage).
Results
We identified 49 eligible studies: 26 operationalized brain health in youth, 13 measured outcomes which they related to brain health and 10 were reviews about youth brain health. Operationalizations of brain health varied widely and included multimodal measures involving questionnaires, cognitive tests, neuroimaging and blood biomarkers. Identified research themes were obesity/fitness/lifestyle as determinants of youth brain health, neonatal brain health, traumatic brain injury and brain health, technological options for measuring brain health, and childhood determinants of brain health.
Conclusion
This review offers a comprehensive overview of possibilities for measuring brain health in youth, which could serve as a valuable foundation for a commonly accepted definition, framework and operationalization of brain health in youth.
{"title":"Brain health in youth - what are we measuring? – A comprehensive review","authors":"Susanne R. De Rooij , Amber Boots","doi":"10.1016/j.braindev.2025.104491","DOIUrl":"10.1016/j.braindev.2025.104491","url":null,"abstract":"<div><h3>Background</h3><div>There is a large body of literature on brain health, usually focusing on brain health in adulthood in relation to neurodegenerative diseases. Over the past ten years, brain health has also emerged as a concept in studies conducted in youth. But what is brain health in youth and how can we measure this?</div></div><div><h3>Methods</h3><div>We summarized the literature on brain health in youth and inventoried operationalizations of brain health and research themes. We searched Medline for studies reporting on brain health in youth (from neonates to late adolescent stage).</div></div><div><h3>Results</h3><div>We identified 49 eligible studies: 26 operationalized brain health in youth, 13 measured outcomes which they related to brain health and 10 were reviews about youth brain health. Operationalizations of brain health varied widely and included multimodal measures involving questionnaires, cognitive tests, neuroimaging and blood biomarkers. Identified research themes were obesity/fitness/lifestyle as determinants of youth brain health, neonatal brain health, traumatic brain injury and brain health, technological options for measuring brain health, and childhood determinants of brain health.</div></div><div><h3>Conclusion</h3><div>This review offers a comprehensive overview of possibilities for measuring brain health in youth, which could serve as a valuable foundation for a commonly accepted definition, framework and operationalization of brain health in youth.</div></div>","PeriodicalId":56137,"journal":{"name":"Brain & Development","volume":"48 1","pages":"Article 104491"},"PeriodicalIF":1.3,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738490","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
To develop a multimodal artificial intelligence (AI) fusion model predicting abnormal fetal brain development from magnetic resonance imaging (MRI).
Methods
Using fetal brain MRI data and clinical indicators from pregnant women (from January 2021 to December 2023), who were split 7:3 into training and validation sets. In the training set, key predictors were identified via univariate analysis and multivariate logistic regression, including both clinical indicators and continuous MRI biometric parameters. Three multimodal AI fusion models，including Convolutional Neural Network-Recurrent Neural Network (CNN-RNN) model, attention mechanism-based model, and feature concatenation model were developed. Performance was assessed by accuracy, precision, recall, F1-score, and the area under the receiver operating characteristic curve (AUC).
Results
Among the total 806 participants, 108 cases (19.15 %) had fetal brain abnormalities in the training set (n = 564), 45 cases (18.59 %) in the validation set (n = 242). Multivariate logistic regression analysis showed that gestational age, gestational diabetes mellitus, alpha-fetoprotein, lateral ventricular width, and sulcation development score were independent risk factors for fetal brain abnormalities. The attention mechanism fusion model achieved the highest AUC in both the training set (0.876) and the validation set (0.869), significantly outperforming the CNN-RNN fusion model (AUC in training set: 0.776; AUC in validation set: 0.718) and the feature concatenation fusion model (AUC in training set: 0.754; AUC in validation set: 0.720).
Conclusion
The multimodal AI fusion model, particularly using attention mechanisms, effectively identifies high-risk fetal brain abnormalities, offering potential for early clinical intervention and improved prenatal counseling to enhance detection and prognosis of neurological disorders.
{"title":"Construction and clinical validation of a fetal brain magnetic resonance imaging-prediction model based on multimodal AI fusion algorithm","authors":"BingGuang Liu, FangJing Zhang, JiMin Guo, Wei Lu, ZhiJun Zhu, Yang Liu, ChenWang Yin","doi":"10.1016/j.braindev.2025.104492","DOIUrl":"10.1016/j.braindev.2025.104492","url":null,"abstract":"<div><h3>Objective</h3><div>To develop a multimodal artificial intelligence (AI) fusion model predicting abnormal fetal brain development from magnetic resonance imaging (MRI).</div></div><div><h3>Methods</h3><div>Using fetal brain MRI data and clinical indicators from pregnant women (from January 2021 to December 2023), who were split 7:3 into training and validation sets. In the training set, key predictors were identified via univariate analysis and multivariate logistic regression, including both clinical indicators and continuous MRI biometric parameters. Three multimodal AI fusion models，including Convolutional Neural Network-Recurrent Neural Network (CNN-RNN) model, attention mechanism-based model, and feature concatenation model were developed. Performance was assessed by accuracy, precision, recall, F1-score, and the area under the receiver operating characteristic curve (AUC).</div></div><div><h3>Results</h3><div>Among the total 806 participants, 108 cases (19.15 %) had fetal brain abnormalities in the training set (<em>n</em> = 564), 45 cases (18.59 %) in the validation set (<em>n</em> = 242). Multivariate logistic regression analysis showed that gestational age, gestational diabetes mellitus, alpha-fetoprotein, lateral ventricular width, and sulcation development score were independent risk factors for fetal brain abnormalities. The attention mechanism fusion model achieved the highest AUC in both the training set (0.876) and the validation set (0.869), significantly outperforming the CNN-RNN fusion model (AUC in training set: 0.776; AUC in validation set: 0.718) and the feature concatenation fusion model (AUC in training set: 0.754; AUC in validation set: 0.720).</div></div><div><h3>Conclusion</h3><div>The multimodal AI fusion model, particularly using attention mechanisms, effectively identifies high-risk fetal brain abnormalities, offering potential for early clinical intervention and improved prenatal counseling to enhance detection and prognosis of neurological disorders.</div></div>","PeriodicalId":56137,"journal":{"name":"Brain & Development","volume":"48 1","pages":"Article 104492"},"PeriodicalIF":1.3,"publicationDate":"2025-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738496","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Paediatric neurology has evolved significantly since its early recognition as a subspecialty in the mid-20th century, though interest in childhood neurological conditions dates back centuries. In Sub-Saharan Africa (SSA), however, the field remains underdeveloped, despite a rising burden of neurological disorders such as epilepsy, cerebral palsy, cerebral malaria, autism spectrum disorder, and paediatric brain tumours. These conditions represent a significant proportion of paediatric morbidity and mortality in the region, yet limited epidemiological data, underdiagnosis, and health system constraints continue to obscure their true impact. Aetiological factors in SSA are diverse and include infectious diseases, perinatal complications, genetic disorders and environmental exposures. While countries such as South Africa, have made strides in diagnosis and care, progress remains uneven across the region. Structured training programmes like the African Paediatric Fellowship Programme and public engagement initiatives have contributed to capacity building, but most countries in the region still lack adequate specialist care, essential diagnostic tools such as electroencephalography and magnetic resonance imaging, and sustained investment in paediatric neurology infrastructure and epidemiological research. To address these gaps, this narrative review recommends expanding local training programmes, integrating task-shifting approaches to empower general practitioners and community health workers, and investing in clinical and epidemiological research. Equally critical is the need to strengthen health systems, improve access to diagnostic services, and promote inclusive, stigma reducing advocacy. Only through coordinated efforts can paediatric neurological care in SSA be advanced to meaningfully improve child health outcomes across the region. Thus, this narrative review explores the evident burdens of paediatric neurology care in SSA and proposes potential strategies to address these challenges.
{"title":"Paediatric neurological care in Sub-Saharan Africa: Current status and future directions","authors":"Mabel Frimpong , Siham Mohamed , Miracle Olayeri Ibukun , Yaa Asieduwaa Owusu , Andrew Awuah Wireko","doi":"10.1016/j.braindev.2025.104490","DOIUrl":"10.1016/j.braindev.2025.104490","url":null,"abstract":"<div><div>Paediatric neurology has evolved significantly since its early recognition as a subspecialty in the mid-20th century, though interest in childhood neurological conditions dates back centuries. In Sub-Saharan Africa (SSA), however, the field remains underdeveloped, despite a rising burden of neurological disorders such as epilepsy, cerebral palsy, cerebral malaria, autism spectrum disorder, and paediatric brain tumours. These conditions represent a significant proportion of paediatric morbidity and mortality in the region, yet limited epidemiological data, underdiagnosis, and health system constraints continue to obscure their true impact. Aetiological factors in SSA are diverse and include infectious diseases, perinatal complications, genetic disorders and environmental exposures. While countries such as South Africa, have made strides in diagnosis and care, progress remains uneven across the region. Structured training programmes like the African Paediatric Fellowship Programme and public engagement initiatives have contributed to capacity building, but most countries in the region still lack adequate specialist care, essential diagnostic tools such as electroencephalography and magnetic resonance imaging, and sustained investment in paediatric neurology infrastructure and epidemiological research. To address these gaps, this narrative review recommends expanding local training programmes, integrating task-shifting approaches to empower general practitioners and community health workers, and investing in clinical and epidemiological research. Equally critical is the need to strengthen health systems, improve access to diagnostic services, and promote inclusive, stigma reducing advocacy. Only through coordinated efforts can paediatric neurological care in SSA be advanced to meaningfully improve child health outcomes across the region. Thus, this narrative review explores the evident burdens of paediatric neurology care in SSA and proposes potential strategies to address these challenges.</div></div>","PeriodicalId":56137,"journal":{"name":"Brain & Development","volume":"48 1","pages":"Article 104490"},"PeriodicalIF":1.3,"publicationDate":"2025-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145738494","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-07DOI: 10.1016/j.braindev.2025.104487
Susanna Stjerna , Lina-Maria Hämäläinen , Mari Videman
Background
Antiseizure medications (ASM) are essential for patients with epilepsy. Though prenatal exposure to ASMs is associated with increased risk for malformations and neurocognitive problems, whether prenatal ASM exposure modifies offsprings' natural developmental trajectory has not yet been studied.
Methods
This prospective study explores the effect of prenatal ASM exposure on the trajectory of cognitive development of children by studying the association of Bayley Scales III scores at 2 years with WISC-IV scores at 6 years of age and by comparing the results against those of unexposed children. Neurocognitive performance of 30 children with prenatal ASM exposure and 37 unexposed control children were evaluated. Correlations and separate ANCOVAs across these ages were compared between ASM exposed and unexposed controls. Results were controlled for maternal education, type of maternal epilepsy, child sex and child age at the assessment.
Results
In unexposed participants, cognitive scores at the age of two years associated positively with working memory and processing speed at six years of age and receptive language scores at the age of two years associated with working memory at six years old. Conversely, with exposed children, there were no significant associations between two- and six-year test scores, and coefficients between receptive language and six-year-old working memory or processing speed differed significantly from unexposed children's coefficients. However, small sample size restricts the stability of the results, and the observed group differences in coefficients were not significant after removal of outlier.
Conclusion
ASM exposure in utero may affect the trajectory of neurocognitive development, but the findings were impacted by an outlier and should be confirmed in larger cohort.
{"title":"Does in utero exposure to antiseizure medications affect the trajectory of cognitive development from 2 to 6 years of age?","authors":"Susanna Stjerna , Lina-Maria Hämäläinen , Mari Videman","doi":"10.1016/j.braindev.2025.104487","DOIUrl":"10.1016/j.braindev.2025.104487","url":null,"abstract":"<div><h3>Background</h3><div>Antiseizure medications (ASM) are essential for patients with epilepsy. Though prenatal exposure to ASMs is associated with increased risk for malformations and neurocognitive problems, whether prenatal ASM exposure modifies offsprings' natural developmental trajectory has not yet been studied.</div></div><div><h3>Methods</h3><div>This prospective study explores the effect of prenatal ASM exposure on the trajectory of cognitive development of children by studying the association of Bayley Scales III scores at 2 years with WISC-IV scores at 6 years of age and by comparing the results against those of unexposed children. Neurocognitive performance of 30 children with prenatal ASM exposure and 37 unexposed control children were evaluated. Correlations and separate ANCOVAs across these ages were compared between ASM exposed and unexposed controls. Results were controlled for maternal education, type of maternal epilepsy, child sex and child age at the assessment.</div></div><div><h3>Results</h3><div>In unexposed participants, cognitive scores at the age of two years associated positively with working memory and processing speed at six years of age and receptive language scores at the age of two years associated with working memory at six years old. Conversely, with exposed children, there were no significant associations between two- and six-year test scores, and coefficients between receptive language and six-year-old working memory or processing speed differed significantly from unexposed children's coefficients. However, small sample size restricts the stability of the results, and the observed group differences in coefficients were not significant after removal of outlier.</div></div><div><h3>Conclusion</h3><div>ASM exposure in utero may affect the trajectory of neurocognitive development, but the findings were impacted by an outlier and should be confirmed in larger cohort.</div></div>","PeriodicalId":56137,"journal":{"name":"Brain & Development","volume":"48 1","pages":"Article 104487"},"PeriodicalIF":1.3,"publicationDate":"2025-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145696507","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号