基于磁共振成像的机器学习在眼眶肿瘤中的应用进展  

Progress of machine learning based on magnetic resonance imaging for orbital tumor research

作　　者：王燕[1] 吴旭莎 胡文鐘 李艳 席一斌 印弘 WANG Yan;WU Xusha;HU Wenzhong;LI Yan;XI Yibin;YIN Hong(College of Life Science,Northwestern University,Xi'an 710068,China;Department of Medical Imaging,Xi'an People's Hospital,Xi'an 710100,China)

机构地区：[1]西北大学生命科学学院,西安710068 [2]西安市人民医院(西安市第四医院)影像科,西安710100

出　　处：《磁共振成像》2024年第8期212-217,共6页Chinese Journal of Magnetic Resonance Imaging

摘　　要：眼眶肿瘤的位置和组织病理学表现各不相同,给诊断带来了挑战。尽管成像技术的进步改善了这一问题,但其分类与鉴别仍是一项挑战。机器学习作为人工智能的关键分支在医疗领域已取得了一定的成果,尤其是其与影像学、眼科学的结合极大地促进了眼眶肿瘤的精准治疗,其在肿瘤鉴别、病灶分割及图像重建等方面已经展现出极大的潜力和广阔的应用前景,有望在提升眼眶肿瘤诊断水平的同时降低临床实践成本。本文就基于MRI的机器学习技术在眼眶肿瘤中的应用进行综述,旨在为临床医师和放射科医生就眼眶肿瘤的诊断、治疗及预后提供思路,并进一步促进机器学习在该领域的应用与普及推广。Orbital tumors vary in location and histopathological differences,presenting diagnostic challenges.Although advances in imaging technology have improved this problem,its classification and identification remains a challenge.As a key branch of artificial intelligence,machine learning has achieved certain results in the medical field,especially its combination with imaging and ophthalmology has greatly promoted the precision treatment of orbital tumors,and it has shown great potential and broad application prospects in tumor identification,lesion segmentation and image reconstruction,which is expected to improve the diagnosis level of orbital tumors and reduce the cost of clinical practice.This article reviews the application of MRI-based machine learning technology in orbital tumors,aiming to provide clinicians and radiologists with ideas for the diagnosis,treatment and prognosis of orbital tumors,and to further promote the application and popularization of machine learning in this field.

关 键 词：眼眶肿瘤 磁共振成像 机器学习 鉴别诊断 疗效预测 预后 

分 类 号：R445.2[医药卫生—影像医学与核医学] R739.7[医药卫生—诊断学]