基于功能核磁影像的精神分裂症患者纹状体功能检测方法

本发明提供了一种定量化方法来检测精神分裂症患者纹状体的健康程度。基于已有的精神分裂症和健康对照的功能核磁共振数据集，提取有关纹状体的复杂功能特征，并通过结合人工智能技术来学习纹状体在精神分裂症中的损伤模式，学习到的模型可泛化到新的精神分裂症患者的功能核磁影像数据，并对该患者的纹状体功能进行定量化评估。 基于该发明得到的纹状体健康程度分数，不仅可在临床提供一定的诊断价值，还可作为生物标记物对精神分裂症患者的用药提供个体化指导。

This invention proposes a quantitative method for evaluating the health status of the striatum in patients with schizophrenia. Based on existing functional magnetic resonance imaging (fMRI) datasets of schizophrenia patients and healthy controls, complex functional features associated with the striatum are extracted. By integrating artificial intelligence technologies, the trained model captures the damage patterns of the striatum in schizophrenia patients, and can be generalized to fMRI data from new schizophrenia patients to quantitatively assess their striatal function. The striatal health score derived from this invention not only offers certain clinical diagnostic value, but also can serve as a biomarker to provide individualized medication guidance for schizophrenia patients.