The invention belongs to the field of psychiatry, nerve images and artificial intelligence, discloses a construction method of a first episode schizophrenia individualized prediction model, and solvesthe problem of low accuracy of auxiliary diagnosis of an existing SCH brain structure network model. The method comprises the following steps: A, acquiring a diffusion tensor image of a first episodeschizophrenia patient; B, preprocessing the obtained diffusion tensor image; C, constructing a sparse brain structure network based on the preprocessed image; D, constructing each sparse multi-threshold fusion brain structure network of the subject by adopting a similar network fusion method; E, extracting multi-threshold fusion brain structure network topology attribute features, and then carrying out feature screening; F, based on the screened features, adopting a classifier for classification training, and obtaining a first episode schizophrenia individualized prediction model is obtained;and G, performing performance verification and evaluation on the first episode schizophrenia individualized prediction model obtained by training.