Department of Otorhinolaryngology-Head and Neck Surgery, Busan Paik Hospital, Inje University College of Medicine, Busan, South Korea.
Department of Radiology, Busan Paik Hospital, Inje University College of Medicine, Busan, South Korea.
PLoS One. 2019 Jul 3;14(7):e0219308. doi: 10.1371/journal.pone.0219308. eCollection 2019.
In this study, we compared ultrasound (US) features between normal parotid parenchyma (NPP) and incidental diffuse parotid disease (DPD). From January 2008 to December 2017, 180 patients underwent neck US before parotid surgery at our hospital. From these, 82 were excluded because of the lack of histopathological data concerning the parotid parenchyma or inadequate US images. A single radiologist blinded to the clinicoserological data and histopathological results, retrospectively investigated all US features and categorizations for the parotid glands using a picture archiving and communication system. Retrospective histopathological analysis of the parotid parenchyma was performed by a single pathologist. On the basis of the histopathological analyses, the 98 patients were divided into NPP (n = 70) and DPD (n = 28) groups. Among US features, parenchymal echogenicity and echotexture showed statistically significant differences between the two groups (p < 0.0001), whereas the gland size, margin, and vascularity showed no significant differences (p > 0.05). The US-based categorization significantly differentiated between NPP and DPD (p < 0.0001), and receiver operating characteristic curve analysis revealed that US categorization based on ≥2 abnormal US features showed the best diagnostic performance for detecting DPD. Thus, US can aid in differentiating DPD from NPP.
在这项研究中,我们比较了正常腮腺实质(NPP)和偶然弥漫性腮腺疾病(DPD)的超声(US)特征。从 2008 年 1 月至 2017 年 12 月,我院 180 例行腮腺手术的患者在手术前接受了颈部 US 检查。其中 82 例因缺乏有关腮腺实质的组织病理学数据或 US 图像不充分而被排除在外。一位对临床血清学数据和组织病理学结果均不知情的放射科医生使用图像存档和通信系统回顾性地研究了所有 US 特征和腮腺分类。由一位病理学家对腮腺实质进行回顾性组织病理学分析。根据组织病理学分析,98 例患者被分为 NPP(n = 70)和 DPD(n = 28)组。在 US 特征中,实质回声和回声纹理在两组之间存在统计学差异(p < 0.0001),而腺体大小、边缘和血管无明显差异(p > 0.05)。基于 US 的分类能显著区分 NPP 和 DPD(p < 0.0001),并且接受者操作特征曲线分析显示,基于≥2 种异常 US 特征的 US 分类对检测 DPD 具有最佳的诊断性能。因此,US 有助于区分 DPD 和 NPP。