Reliability of imaging based diagnosis of lateral ventricular masses in children.

WORLD NEUROSURGERY(2019)

引用 0|浏览11
暂无评分
摘要
OBJECTIVE: We studied the accuracy of the radiologic diagnosis of lateral ventricular masses in children (<20 years of age). METHODS: In this retrospective study, data were collected from children with lateral ventricular masses managed in our unit between 2001 and 2016. There were 43 patients (26 boys and 17 girls; mean age, 12.1 years; range, 8 months to 20 years). Magnetic resonance imaging was available for 36 patients, whereas in 7 patients only a contrast-enhanced computed tomography scan was available. The images were read independently by 3 radiologists, who were blinded to the pathology. Two differential diagnoses were offered for each patient. The agreement between the 3 radiologists was calculated using the Fleiss kappa statistic. RESULTS: The common pathologic diagnoses were subependymal giant cell astrocytoma (SEGA) (20.9%), low-grade astrocytoma (16.3%), high-grade astrocytoma (9.3%), choroid plexus papilloma (9.3%), and meningioma (9.3%). The sensitivity, specificity, and accuracy of the radiologic diagnoses were 62%, 96.7%, and 94.2%, respectively. Low-grade tumors such as low-grade gliomas, SEGAs, choroid plexus papillomas, and meningiomas were diagnosed with a high level of accuracy. High-grade gliomas, choroid plexus carcinomas, and other malignant neoplasms were difficult to diagnose on imaging. Cavernous angiomas were also difficult to diagnose. There was only fair agreement between the 3 radiologists (Fleiss kappa = 0.24). CONCLUSIONS: The pathologic spectrum of lateral ventricular tumors in children is wide, and identifying the pathology on imaging is difficult for malignant tumors. Benign tumors such as SEGAs, low-grade astrocytomas, and choroid plexus papillomas are relatively easier to diagnose. There is significant interobserver variability in the radiologic diagnosis of these tumors.
更多
查看译文
关键词
Brain tumor,Children,Diagnosis,Lateral ventricle,Radiology
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要