人工智能在口腔癌和口腔潜在恶性疾病诊断中的研究进展  

Research progress on artificial intelligence techniques in diagnosis of oral cancer and potentially malignant disorders

作　　者：林慧平[1] 徐婷[1] 林军[1] Lin Huiping;Xu Ting;Lin Jun(Dept.of Stomatology,The First Affiliated Hospital,Zhejiang University School of Medicine,Hangzhou 310003,China)

机构地区：[1]浙江大学医学院附属第一医院口腔科,杭州310003

出　　处：《国际口腔医学杂志》2023年第2期138-145,共8页International Journal of Stomatology

基　　金：国家自然科学基金(81970978);浙江省自然科学基金(LQ-20H140007)。

摘　　要：口腔癌是一个相当普遍的全球健康问题,发病率和死亡率逐年上升。加强对口腔癌的早发现、早诊断,特别是防控在口腔潜在恶性疾病(OPMD)阶段,是实现降低口腔癌发病率和死亡率的关键。然而,口腔癌及OPMD的筛查和早期诊断是一项艰巨的医学任务,缺乏简便易行的诊断方法。基于人工智能的口腔医学图像的检测和诊断技术为此提供了新的思路和方法。本文将从组织病理学图像、共聚焦激光显微内镜图像、高光谱成像、自体荧光成像以及自然光图像5种口腔图像着手,解析目前人工智能在口腔癌和OPMD检测诊断中的研究进展和存在的困难。Oral cancer is a quite common global health issue, with morbidity and mortality progressively increased in recent years. Enhancing early diagnosis of oral cancer, especially in the phase of oral potentially malignant disorders(OPMD), would significantly reduce the incidence and mortality of oral cancer. However, earlier diagnosis of oral cancer and OPMD is still very challenging, and lacks of effective and efficient(traditional) medical methods. Artificial intelligence(AI) provides new insights and approaches for diagnosis of oral disease. Here we review the applications of AI to detect oral cancer and OPMD on a variety of photographic images, including histopathological images, confocal laser endomicroscopy, hyperspectral imaging, and white light images. And then, the obstacles for AI in diagnosis are also given.

关 键 词：口腔癌 口腔潜在恶性疾病 人工智能 

分 类 号：R782[医药卫生—口腔医学]