多模态超声与其联合深度学习在乳腺癌诊断中的研究进展  

Research progress of multimodal ultrasound and its combination with deep learning in breast cancer diagnosis

作　　者：王琪 宋宏萍 许磊 WANG Qi;SONG Hongping;XU Lei(School of Medical Technology,Shaanxi University of Chinese Medicine,Xianyang 712046,China;Department of Ultrasound,Xijing Hospital,Air Force Military Medical University,Xi'an 710032,China;Department of Functional,Xi'an Hospital of Traditional Chinese Medicine,Xi'an 710021,China)

机构地区：[1]陕西中医药大学医学技术学院,陕西咸阳712046 [2]空军军医大学西京医院超声医学科,陕西西安710032 [3]西安市中医医院功能科,陕西西安710021

出　　处：《分子影像学杂志》2024年第10期1119-1123,共5页Journal of Molecular Imaging

基　　金：国家自然科学基金面上项目(82471991,82071934);陕西省国际科技合作与交流计划重点项目(2020KWZ-022)。

摘　　要：乳腺癌是导致女性死亡的重要原因。超声是乳腺癌疾病的主要影像学检查方法,将二维超声、彩色多普勒血流成像、超声弹性成像等不同超声技术联合应用于乳腺癌的诊断可显著提高诊断效率和准确性,减少不必要的穿刺活检。人工智能技术的加入,有潜力帮助医生更高效、更精准的作出决策,为诊断乳腺癌提供一种新的策略。本文将对比不同超声技术在乳腺癌诊断中的优劣,讨论深度学习联合多模态超声成像在乳腺癌诊断、预测和疗效评估等方面的积极作用,并提出乳腺超声未来可能面临的挑战,以期为临床医生提供参考。Breast cancer is an important cause of death in women.Ultrasound is the main imaging examination method for breast cancer.The combined application of different ultrasound technologies such as two-dimensional ultrasound,color Doppler flow imaging,and ultrasound elastography to the diagnosis of breast cancer can significantly improve diagnostic efficiency and accuracy,and reduce unnecessary needle biopsy.The addition of artificial intelligence technology has the potential to help doctors make more efficient and accurate decisions,providing a new strategy for diagnosing breast cancer.This review compared the advantages and disadvantages of different ultrasound technologies in the diagnosis of breast cancer,discussed the positive role of deep learning combined with multi-modal ultrasound imaging in the diagnosis,prediction and efficacy evaluation of breast cancer,and proposed the challenges that breast ultrasound may face in the future,with a view to providing clinical doctors with reference.

关 键 词：乳腺癌 超声 诊断 深度学习 

分 类 号：R737.9[医药卫生—肿瘤] R445.1[医药卫生—临床医学]