深度学习模型在侵袭性垂体腺瘤诊疗中的应用

Application of deep learning model in the diagnosis and management of invasive pituitary adenomas

垂体腺瘤通常为良性肿瘤,但约1/3的病变会侵袭肿瘤周围的正常组织。Knosp分级和Hardy分级作为临床工作中评估垂体腺瘤侵袭性的主要方法,存在明显的局限性。深度学习属于人工智能机器学习的热门领域,是一种基于人工神经网络的新兴技术。深度学习模型能够自动分析垂体腺瘤的术前影像数据,提高诊断肿瘤侵袭性的准确率,可以帮助临床医生更好地拟定手术入路和切除方式,在临床诊疗中有巨大的应用前景。

Pituitary adenomas are generally considered benign tumors,but approximately one third of the lesions invade the surrounding normal tissues.As the main methods to evaluate the invasiveness of pituitary adenomas in clinical practice,Knosp grades and Hardy grades have significant limitations.Deep learning belongs to the hot field of machine learning of artificial intelligence,which is an emerging technology based on artificial neural network.The deep learning model can automatically analyze the preoperative image data of pituitary adenoma,improve the accuracy of diagnosis of tumor invasiveness,and help to better determine the surgical approach and resection methods,which has great application prospects in clinical diagnosis and treatment.