Hengyang Lu ¹ Heng Zhang ² Yi Zhong ¹ Xiang-Yu Meng ¹ Meng-Fei Zhang ¹ Qiu Ting ^2*

• ¹ Jiangnan University, Wuxi, China
• ² Wuxi Maternity and Child Health Care Hospital, Wuxi, Jiangsu Province, China

Autism spectrum disorder (ASD) is a neurodevelopmental condition that significantly impacts the mental, emotional, and social development of children. Early screening for ASD typically involves the use of a series of questionnaires. With answers to questionnaires, healthcare professionals can identify whether a child is at risk for developing ASD and refer them for further evaluation and diagnosis. CHAT-23 is an effective and widely used screening test in China for the early screening of ASD, which contains 23 different kinds of questions. Machine learning can learn experience from known data and predict unknown data. Thus, we propose a machine learning based method for ASD early screening from the perspective of computer-aided diagnosis. We have collected the clinical data from Wuxi, China. We regard all the questions of CHAT-23 as different kinds of features for building machine learning models. On the one hand, we introduce machine learning methods into ASD, providing accurate and more effective tests for early screening for autism. On the other hand, we use the most Max-Relevance and Min-Redundancy (mRMR) feature selection method to analyze the most important questions among all 23 from the collected CHAT-23 questionnaires. We build seven mainstream supervised machine learning models and conduct experiments. Our study mainly focused on the health of Chinese children.