This study aims to address the mental health challenges brought about by the diversified development and rapid changes in society, with special attention to the psychological status of the student population. By using the SCL-90 mental health testing tool, collecting students' mental health data, and applying the fuzzy comprehensive evaluation method to analyze and evaluate students' mental health and its influencing factors in depth, the study aims to provide more effective countermeasures for students' mental health education as well as targeted teaching assistance for teachers. This study combines the BP neural network prediction model, which is committed to improving the accurate prediction of students' mental health status. The results of the study will help to assess the mental health level of students, detect and intervene in psychological crises in a timely manner, provide schools with more comprehensive mental health management and services, and promote the overall healthy growth of students.