医学
回声
门(解剖学)
同种类的
淋巴
接收机工作特性
放射科
置信区间
淋巴结
病理
单变量分析
颈淋巴结
人口
超声波
核医学
内科学
多元分析
环境卫生
癌症
热力学
物理
转移
作者
Sun-Young Park,Ji Young Kim,Young Jin Ryu,Hyunju Lee
摘要
Objectives To investigate ultrasound (US) features of enlarged cervical lymph nodes (LNs) to differentiate between Kikuchi disease (KD) and other common types of infectious lymphadenitis in an East Asian pediatric patient population. Methods A total of 142 pediatric patients with KD and 45 patients with infectious lymphadenitis (suppurative lymphadenitis [n = 29], nontuberculous mycobacterial lymphadenitis [n = 9], and tuberculous lymphadenitis [n = 7]) were included. The clinical characteristics, laboratory results, and US features of LNs were reviewed. The area under the curve (AUC) from a receiver operating characteristic curve analysis was used as a diagnostic accuracy measure. Results A multiple clustered adjacent pattern, bilaterality, an even size, posterior neck involvement, no enlargement, an elongated‐to‐ovoid shape, homogeneous hypoechogenicity, a well‐defined margin, presence of an echogenic fatty hilum, no intranodal gross necrosis, increased perinodal fat echogenicity, and no increased echogenicity of the adjacent sternocleidomastoid muscle were significant US features of the affected LNs to discriminate KD from infectious lymphadenitis ( P < .05). Homogeneous hypoechogenicity in KD showed the highest AUC (0.930) as a single variable (95% confidence interval, 0.88–0.96). The AUCs were increased in 3 combination models with 2 US features: homogeneous echogenicity and 1 of 3 other US features (increased perinodal fat echogenicity, 0.935; number of affected LNs, 0.947; and LN shape, 0.949). Conclusions Homogeneous hypoechogenicity of LNs was a significant US feature with the highest diagnostic accuracy in differentiating KD from common infectious lymphadenitis on a univariate analysis. In the combination model, US features of an elongated‐to‐ovoid shape and homogeneous hypoechogenicity showed the highest diagnostic accuracy.
科研通智能强力驱动
Strongly Powered by AbleSci AI