Early Screening is vital and helpful for implementing intensive intervention and rehabilitation therapy for the children with Autism Spectrum Disorder (ASD). Research has shown that Electroencephalogram (EEG) signals can reflect abnormal brain function of the children with ASD and screening with EEG signals has the characteristics of good real-time performance and high sensitivity. However, the existing EEG screening algorithms mostly focus on the data analysis in the resting state and the extracted EEG features have some drawbacks such as weak representation capacity and information redundancy. In this work, we utilize Event-related Potential (ERP) technique to acquire testees’ EEG data under positive and negative emotional stimulation and propose a EEG Feature Selection Algorithm based on L1-norm Regularization to carry out screening of autism. The proposed EEG Feature Selection Algorithm include the following steps: 1) Extracting 20 EEG features from raw data; 2) Classification with Support Vector Machine (SVM); 3) Selecting appropriate EEG feature with L1- norm Regularization according to the classification performance. The experimental results shows that the accuracy for screening of the children with ASD can reach 93.8% and 87.5% under positive and negative emotional stimulation and the proposed algorithm can effectively eliminate redundant features and improve screening accuracy.

中文翻译：

---

基于L1-范式正则化的脑电信号特征选择对ASD儿童进行早期筛查

早期筛查对于对自闭症谱系障碍（ASD）的儿童实施强化干预和康复治疗至关重要，并且很有帮助。研究表明，脑电图（EEG）信号可以反映ASD儿童的脑功能异常，并且用EEG信号进行筛查具有实时性好和敏感性高的特点。然而，现有的脑电图筛选算法大多侧重于静止状态下的数据分析，提取的脑电图特征具有表示能力弱，信息冗余等缺点。在这项工作中，我们利用事件相关电位（ERP）技术在正向和负向情绪刺激下获取睾丸的脑电数据，并提出基于L1范数正则化的脑电特征选择算法来进行自闭症筛查。提出的脑电特征选择算法包括以下步骤：1）从原始数据中提取20个脑电特征。2）使用支持向量机（SVM）进行分类；3）根据分类性能，通过L1-范数正则化选择合适的EEG特征。实验结果表明，在积极和消极的情绪刺激下，ASD儿童的筛查准确率可分别达到93.8％和87.5％，所提算法可以有效消除多余特征，提高筛查准确性。