面向儿童脑发育障碍性疾病的神经机制建模与辅助诊疗算法

Epilepsy and autism are two of the most common neurodevelopmental disorders in children. The study of these two diseases is of great significance for the early identification and treatment of neurodevelopmental disorders in children. Despite the amount of pioneer work that has been done, existing studies have many limitations due to the strong heterogeneity of clinical symptoms, complex causes and poor data accuracy. This project aims to improve the existing methods by (1) innovatively exploring the electrophysiological neural mechanism of childhood epilepsy at three different scales, namely, neurons, neuronal groups and brain regions, and hence provides possible new therapeutic targets for the treatment of childhood epilepsy; (2) developing deep neural network models to improve the accuracy and robustness of epileptic seizure detection and early diagnosis of autism; (3) building statistical models to predict the onset time of seizure; (4) selecting optimal individualized treatment regimes; (5) providing clinicians with an Artificial Intelligent (AI) diagnostic report and individualized treatment plan for clinical application and evaluation. Overall, we aim to develop new theories, methods and ideas for the study of epilepsy and autism, establish an AI analysis system and a diagnostic decision support system for childhood epilepsy and autism, so as to promote the research and application in the field of neurodevelopmental disorders.

癫痫和自闭症是儿童脑发育障碍性疾病中最具有代表性的两种疾病，对此两种疾病的研究对于儿童脑发育障碍性疾病早期识别、治疗具有重要意义。尽管学术各界已开展大量工作，但由于癫痫和自闭症临床表型异质性强、病因复杂、数据精度差等原因，现有的方法仍存在众多缺陷。本项目将改进现有方法，创新性地在神经元、神经元群和脑区三个不同尺度上探究儿童癫痫发作时的电生理神经机制，为儿童癫痫治疗提供可能的新治疗靶点；再通过电生理数据以及患儿的各项可监测指标，建立深度神经网络模型以提升癫痫发作检测/自闭症早期诊断的准确性和稳定性，以及建立统计模型以预测癫痫发作时间；根据患儿特征选取精准的个性化治疗方案；为临床医生提供智能诊断报告和精准医疗方案并进行临床应用及评价。本项目将提出新理论、新方法、新思想，建立面向儿童癫痫和自闭症的智能分析系统以及辅助诊断决策支持系统，推动儿童脑发育障碍性疾病的研究和应用向前发展。

癫痫和自闭症是儿童脑发育障碍性疾病中最具有代表性的两种疾病，对此两种疾病的研究对于儿童脑发育障碍性疾病早期识别、治疗具有重要意义。尽管学术各界已开展大量工作，但由于癫痫和自闭症临床表型异质性强、病因复杂、数据精度差等原因，现有的方法仍存在众多缺陷。本项目将改进现有方法，创新性地在神经元、神经元群和脑区三个不同尺度上探究儿童癫痫发作时的电生理神经机制，为儿童癫痫治疗提供可能的新治疗靶点；再通过电生理数据以及患儿的各项可监测指标，建立深度神经网络模型以提升癫痫发作检测/自闭症早期诊断的准确性和稳定性，以及建立统计模型以预测癫痫发作时间；根据患儿特征选取精准的个性化治疗方案；为临床医生提供智能诊断报告和精准医疗方案并进行临床应用及评价。本项目将提出新理论、新方法、新思想，建立面向儿童癫痫和自闭症的智能分析系统以及辅助诊断决策支持系统，推动儿童脑发育障碍性疾病的研究和应用向前发展。