摘要
Abstract
Pituitary adenomas(also known as pituitary neuroendocrine tumors)are one of the relatively common primary brain tumors,and early diagnosis along with appropriate treatment strategies are crucial for patient prognosis.Radiomics has been widely applied in the study of pituitary adenomas,encompassing consistency prediction,invasiveness prediction,histopathological feature prediction,as well as prediction of surgical and non-surgical treatment responses.For consistency prediction,numerous studies based on different magnetic resonance imaging(MRI)sequences and machine learning algorithms,such as support vector machine(SVM)and random forest(RF),have demonstrated good diagnostic performance.In terms of invasiveness prediction,radiomic features can be used to assess tumor invasiveness and histopathological markers.Radiomics also facilitates the differential diagnosis between pituitary adenomas and other pituitary lesions by predicting histopathological features and identifying various pathological subtypes.Regarding treatment response,radiomics can help predict tumor recurrence and complications after surgery and evaluate responses to pharmacological therapy.In the era of precision medicine,radiomics based on machine learning is expected to become increasingly sophisticated and play a more effective role in the accurate diagnosis and individualized treatment of pituitary adenomas and other tumors.关键词
影像组学/机器学习/垂体腺瘤/分子分型/人工智能Key words
radiomics/machine learning/pituitary adenoma/molecular typing/artificial intelligence
