Combining physiological signal analysis as a multidimensional screening system for early detection of infants with ASD

• Main Authors: YU-AN CHEN, 陳育安
• Other Authors: Yuh-Show Tsai
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/a54y32

碩士 === 中原大學 === 生物醫學工程研究所 === 107 === ASD (Autism spectrum disorder) is a congenital defect of brain that causes developmental disorder. If ASD can be detected during infancy and provide appropriate treatment as early as possible, the behavior of incompatibility and devastation of infants will be reduced. Currently, questionnaires survey and Interview with parents are major methods for early detection such as Infant-Toddler Checklist (ITC). These methods are very subjective and might be easily influenced by caregivers. How to objectively screen ASD and provide early intervention are important issues. Prior studies have facilitated Still Face Experiment for infant ASD screening. It is found that infants exhibited calm response during the experiment might be considered as high risk group of ASD. The objective of this study is to quantify the infant’ during Still Face Experiment via Heart rate Variability (HRV), and hence, provide objectively screening apparatus for infants with possible ASD. In this study, an early stage ASD screening system is developed. The infants’ Electrocardiography (ECG) were collected during the Still Face Experiment. The R-R intervals of ECG were extracted and processed by Continuous Wavelet Transform. The transient response of HRV was derived between different time frames. The spectral change along with temporal information can effectively reflect infants’ emotional response. Both simulated ECG signal and Tilt Table Test have been utilized to verify our system functions. Pilot validation of the screening system has been proceeded with 6 infants, aged from 6 to 12 months. By comparing the results with ITC, the high-risk ASD infant group presented relatively low activity of autonomous nerve response. The feasibility and efficacy of our early stage ASD screening are proved.