1 Department of Radiology, Chung-Ang University Hospital, Chung-Ang University College of Medicine, Seoul, South Korea.
2 Department of Radiology, Busan Paik Hospital, Inje University College of Medicine, 75 Bokji-ro, Busanjin-gu, Busan, South Korea 47392.
AJR Am J Roentgenol. 2018 Sep;211(3):649-654. doi: 10.2214/AJR.17.19164. Epub 2018 Jul 11.
The purpose of this multicenter study was to assess the diagnostic accuracy of real-time sonography (US) for differentiating diffuse thyroid disease (DTD) from normal thyroid parenchyma among radiologists blinded to patients' clinical, serologic, and imaging history and to determine the differences in diagnostic accuracy among radiologists from different institutions.
From January to March 2017, 214 patients underwent preoperative thyroid US and subsequent thyroid surgery at four participating institutions. Real-time US was performed at each institution by an attending radiologist, who classified US diagnoses into one of the following four categories based on US findings: no DTD, indeterminate, suspicious for DTD, and DTD. The outcomes of US diagnoses were compared with histopathologic results to determine the diagnostic accuracy of real-time US at each institution.
Histopathologic results included normal thyroid parenchyma (n = 143), Hashimoto thyroiditis (n = 29), non-Hashimoto lymphocytic thyroiditis (n = 37), and diffuse hyperplasia (n = 5). Normal thyroid parenchyma and DTD exhibited statistically significant differences in echogenicity, echotexture, size, glandular margin, vascularity of thyroid, and US classification. There was positive correlation between US classification and histopathologic results at all institutions for detecting DTD. The highest diagnostic indexes were obtained when the cutoff criterion was suspicious for DTD. There was favorable diagnostic accuracy, with statistically significant differences, at all institutions for the diagnosis of DTD.
Real-time US can be helpful for differentiating DTD from normal thyroid parenchyma.
本多中心研究旨在评估超声（US）实时成像在对临床、血清学和影像学资料未知的放射科医师中对弥漫性甲状腺疾病（DTD）与正常甲状腺实质进行鉴别诊断的准确性，并确定不同机构的放射科医师在诊断准确性方面的差异。
2017 年 1 月至 3 月，214 例患者在 4 个参与机构中接受了术前甲状腺 US 和随后的甲状腺手术。在每个机构中，均由一位主治放射科医师进行实时 US，根据 US 结果将 US 诊断分为以下四个类别之一：无 DTD、不确定、疑似 DTD 和 DTD。比较 US 诊断结果与组织病理学结果，以确定各机构实时 US 的诊断准确性。
组织病理学结果包括正常甲状腺实质（n = 143）、桥本甲状腺炎（n = 29）、非桥本淋巴细胞性甲状腺炎（n = 37）和弥漫性增生（n = 5）。正常甲状腺实质和 DTD 在回声、回声纹理、大小、腺体边缘、甲状腺血管和 US 分类方面存在统计学显著差异。在所有机构中，US 分类与组织病理学结果之间存在正相关，用于检测 DTD。当截断标准为疑似 DTD 时，获得了最高的诊断指标。在所有机构中，DTD 的诊断准确性均较高，且具有统计学显著差异。
实时 US 有助于鉴别 DTD 与正常甲状腺实质。
