计算机科学
医学
人工智能
自然语言处理
甲状腺结节
放射科
算法
甲状腺
内科学
作者
Shelly Soffer,Benjamin S. Glicksberg,Eyal Zimlichman,Eyal Klang
标识
DOI:10.1016/j.acra.2021.03.036
摘要
We have read with great interest the recently published article entitled "Management of Incidental Thyroid Nodules on Chest CT: Using Natural Language Processing to Assess White Paper Adherence and Track Patient Outcomes" ( 1 Short R.G. Dondlinger S. Wildman-Tobriner B Management of incidental thyroid nodules on chest CT: using natural language processing to assess white paper adherence and track patient outcomes. Acad Radiol. 2021; (S1076-6332(21)00090-8) Google Scholar ). In this article, Short et al. used a natural language processing (NLP) model to identify incidental thyroid nodules meeting criteria for sonographic follow-up. Their algorithm employed the fastText architecture, developed by Facebook in 2017, for word embedding, and a fully connected deep learning network for classification. Their model results were notable, with a sensitivity of 92.1%, and a specificity of 96.6%.
科研通智能强力驱动
Strongly Powered by AbleSci AI