Muti-modality MRI Study On Neurodevelopment In Rolandic Epilepsy

Objectives:Rolandic epilepsy(RE)is the most common idiopathic childhood epilepsy.Neurodevelopmental abnormalities are considered to be the cause of RE,and it is also the result of epilepsy damage,which is the pathological basis of brain cognitive impairment in children.The study of neurodevelopmental abnormalities in rolandic epilepsy is helpful to understand its neurophysiological mechanism,and can also provide methods and evidence for clinical evaluation of brain cognitive impairment in children.In this study,we will use multimodal MRI to study the development of re from two aspects of brain function and brain structure.Methods:In brain functional development,this study intends to develop and verify the research path of brain functional development based on brain functional gradient by using the method of brain functional diffusion embedding,and use it to explore the changes of multi-level hierarchy development of Rolandic epilepsy large-scale brain network,and use the spatial pattern of postmortem human brain gene transcriptome to verify the potential physiological mechanism behind the specific changes.In brain structure development,this study uses neuroanatomical features from large sample data to predict brain age based on deep learning,and uses the difference between individual predicted brain age and chronological age:brain predicted age difference(PAD)can converge the complex development pattern of the whole brain to a single output.This index was used to verify and quantify the brain development delay of Rolandic epilepsy patients,and to further analyze the relationship between individual brain development delay and clinical symptoms and cognitive impairment.Results:The gradient of brain functional network,especially the eccentricity of brain functional gradient,showed significant changes with age in children and adolescents.In Rolandic epilepsy patients,we found that visual network and dorsal attention network showed the most significant difference in brain gradient and brain gradient eccentricity.The spatial relationship between the spatial pattern of gene transcriptome and the developmental difference of brain gradient eccentricity was analyzed.We found that the transcription of genes related to the developmental difference of brain gradient eccentricity had a developmental enrichment consistent with the brain region and specific age involving by Rolandic epilepsy.It is related to neuron differentiation and development in biological process,synapse in cell composition,epilepsy,and developmental disorders in diseases.Finally,we found the specific contributions of excitatory neurons,inhibitory neurons,astrocytes and microglia.Based on a large sample of structural MRI data and brain age prediction model with novel learning algorism of BAENET,we comprehensively estimated the developmental changes of neuroanatomic features during disease course of Rolandic epilepsy.Rolandic epilepsy patients showed imaging phenotypes of delayed brain development with increased GM volume and decreased WM volume in the Rolandic regions.Strikingly,we found Rolandic epilepsy patients had a 0.45-year delay of brain-PAD by comparing local TDC.We further found that in patients with Rolandic epilepsy,attention has a positive contribution to brain-Pad,and antiepileptic drugs may improve the delay of brain-PAD.Conclusions:In brain functional development,we have not only established a new path and framework for the study of large-scale brain gradient development,but also provided new evidence for abnormal brain functional development of Rolandic epilepsy,bridging the gap between transcriptome and neuroimaging,and promoting the comprehensive understanding of the neural mechanism behind large-scale brain gradient development and abnormal brain functional development in Rolandic epilepsy.In terms of brain structure development,our study confirmed 0.45 years brain age development delay in Rolandic epilepsy patients,which shows that deep learning can accurately predict the brain age of children,and it is a potential biomarker for quantifying the brain development of children with neuropsychiatric diseases.