| Autism,also known as Autism Spectrum Disorder(ASD),is one of the widely prevalent neurodevelopmental disorders.In recent years,a number of studies have explored the structural and functional characteristics of the brains of autistic individuals using magnetic resonance imaging combined with clinical diagnostic symptoms,and found certain structural abnormalties in autistic populations compared to typical developing control(TDC)individuals.These structural abnormalities have strong correlations with abnormalities or even deficiencies of corresponding brain functions.Meanwhile,autism has a complex etiology and can be divided into several social subtypes as a spectrum disorder.Childhood is a critical period for brain development,and it has been increasingly established that there is developmental trajectory abnormality in autism.However,as far as we know,there are few studies which have systematically unveiled the fine detail of structural characteristics differences in young ASD children(especially those aged 4-6 years),and explored the brain structural characteristics in the different social subtypes of ASD.Diffusion Tensor Imaging(DTI),a form of magnetic resonance imaging,allows better visualization of the fine detail of brain white matter tracts as well as the potential structural abnormalities in ASD patients than other imaging modalities.Therefore,in this study,based on the fractional anisotropy(FA)index of brain DTI data,we performed three analysises to systematically investigate 1)the structural connectivity differences between 4-6 year old ASD and TDC,2)the major white matter fiber tracts properties differences between 4-6 year old ASD and TDC,and 3)the correlations between the identified structural connectivity and symptom severity in three social subtypes of ASD using the Autism Diagnostic Observation Schedule(ADOS).Our aim is to identify reliable brain imaging markers for early diagnosis and follow-up treatment of ASD children and their social subtypes.Three independent datasets were adopted in this study to verify the robustness of the results(dataset 1: 95 ASD and 26 TDC children aged 4-6years provided by our collaborator,Associate Professor Rong Zhang’s team at Peking University;dataset 2: 12 ASD children collected at UESTC;dataset 3: 9 TDC children provided by Professor Xiaoyan Ke’s team at Nanjing Brain Hospital).The first study identified five structural brain networks with 22 links showing increased FA values in ASD children compared to TDC,while no decreased ones.The identified five structural brain networks were found to be associated with major functional deficits in ASD such as default mode,motor,visual recognition,morality and reward based on the meta-analysis using neurosynth.The averaged FA values of the five networks were also found to be significantly correlated with ADOS total scores and certain subscores.Based on the 22 links within the five structural brain networks showing significant group differences,we further developed a classification model based on Support Vector Machine(SVM),and achieved more than 90% classification accuracies between ASD and TDC using leave-one-out cross-validation(LOOCV)and independent testing datasets.The second study identified alterations(uncorrected)in specific regions of identified fiber tracts between ASD and TDC using TBSS.The third study identified differential correlations between structural connectivity and certain ADOS scores within the three social subtypes of ASD children.In summary,this study unveiled the differences of 4-6 year old ASD children and their social subtypes compared to TDC children in terms of structural connectivity and white matter fiber tracts properties.The identified brain structural connectivity properties could not only serve as potential brain imaging markers for young ASD children,but also provide novel insights and effective explorations for further early diagnosis of autism based on MRI imaging and artificial intelligence,as well as for precise intervention treatment. |
PDF Full Text Request