Thyroid anomaly identification method and system

According to the thyroid anomaly identification method and system provided by the invention, the potential feature expression is obtained through the thyroid description feature expression corresponding to the paradigm thyroid description data and the anomaly description feature expression, then the to-be-trained identification unit is iteratively trained through the potential feature expression, and the target identification unit is obtained through training. Therefore, the target identification unit can generate the target thyroid description data through the input identification data, that is, multiple thyroid description data can be quickly generated through the target identification unit obtained through training according to requirements, and thus the accuracy and reliability of thyroid anomaly identification are improved. 本申请提供的一种甲状腺异常识别方法及系统，通过范例甲状腺描述数据对应的甲状腺描述特征表示以及异常描述特征表示得到潜在特征表达，然后通过该潜在特征表达对待训练识别单元进行迭代训练，训练得到目标识别单元，该目标识别单元即可通过输入的识别数据生成目标甲状腺描述数据，即，可以根据需求通过训练得到的目标识别单元快速生成多个甲状腺描述数据，从而提升甲状腺异常识别的准确性和可靠