Research On The Analysis Methods Of Fnirs Signals In Adhd Children And Development Of FNIRS System

In China,lots of children suffer from varying degrees of brain function dysplasia and cognitive dysfunction,which influence the children's learning and social abilities seriously.Therefore,early diagnosis and intervention of the brain function dysplasia and cognitive dysfunction were the social problems to be solved urgently for our country.Functional near infrared spectroscopy(fNIRS)as a non-invasive optical imaging technology is suitable to be used in the pediatric population with the advantages of high ecological validity,portablility and continuous measurement.However,the analysis methods of fNIRS obviously lag behind the analysis methods of functional magnetic resonance imaging(fMRI)and electroencephalography(EEG),and the domestic fNIRS devices mainly depend on the imports.To promote the fNIRS research in children with brain function dysplasia and cognitive dysfunction,especially in the diagnosis of children with Attentiondeficit/hyperactivity disorder(ADHD),this study mainly focuses on the development and improvement of analysis methods,the development of advanced fNIRS device and the research of brain intervention technique.Firstly,a motion correction method based on the empirical mode decomposition(EMD)algorithm was proposed,according to the nonlinearity and nonstationarity of fNIRS signal.This method takes full advantage of the adaptivity of EMD to separate the motion artifact signal and the physiological signal effectively.After compared with three other common motion correction methods,the results indicated that this method can decrease the mean square error and improve the signal-noise-ratio.This motion correction method based on EMD is stable and effective.Secondly,permutation entropy was used to mearsure the complexity of fNIRS signal.Permutation entropy is computationally efficient and less sensitive to the data length.It was aimed to investigate the different complexity between the children with and without ADHD,and the relationship between the complexity and the cortical activation.The study found that the complexity exhibited significantly negative correlation with the cortical activations and positive correlation with ADHD index.In addition,the complexity of right dorsolateral prefrontal cortex in ADHD children were significantly larger than those in normal controls.The results suggest that the complexity of fNIRS signal is a potential biomarker of ADHD.Thirdly,based on the spatial pattern of brain activity,two multivariate analysis methods of fNIRS signal were proposed: multivariate pattern analysis and partial least squares.The multivariate pattern analysis discriminated the ADHD children from healthy controls with an overall accuracy of 93.6%,and found that the brain activities in bilateral dorsolateral prefrontal cortex,inferior medial prefrontal cortex,right posterior prefrontal cortex and right temporal cortex were altered in ADHD children compared with controls.The partial least squares found that the brain activities in bilateral dorsolateral prefrontal cortex,partial ventrolateral prefrontal cortex and right temporal cortex were changed with the working memory load.In 1-back condition,the brain activities in these regions were greater in controls than ADHD children.In addition,the dorsolateral prefrontal cortex and right temporal cortex exhibited greater activities in faster and more accurate subjects.The results show the potential feasibility of multivariate analysis methods of fNIRS signal for the early diagnosis of ADHD children.Lastly,according to the current situation of the domestic existing fNIRS system mainly depending on the imported products,a low cost and portable f NIRS system was designed.The corresponding collection software and analysis software were developed.The common analysis methods and the methods proposed in this study were integrated into the analysis software.Based on the f NIRS system,a neurofeedback system for the children was also developed.After testing,the performances of the fNIRS system developed in this study were similar to the commercial NIRScout system.