The invention relates to the field of neurodevelopmental disorders, and discloses an electro-oculogram
signal analysis and diagnosis
system based on neurodevelopmental disorders, which extracts
signal features from pre-processed electro-oculogram signals, and analyzes inter-
group differences between
signal features of a child
patient group and signal features of a healthy child group. The method comprises the following steps: preliminarily screening out signal features with significant
statistical difference between two groups, then carrying out severity
correlation analysis on the screened signal features, screening and forming a feature
list highly related to ADHD severity, and then comparing
eye movement features of a child
patient group under natural behaviors with
eye movement features of a healthy child group, so as to obtain a feature
list related to ADHD severity. The method comprises the following steps of: identifying and forming an abnormal signal feature
list, fusing the feature list highly related to the ADHD severity and the abnormal signal feature list, calculating a comprehensive abnormal
score, drawing an ROC curve according to the calculated comprehensive abnormal
score, and carrying out grading evaluation on the ADHD severity of a child patient, so that accurate diagnosis and severity judgment on the ADHD
disease are realized.