Interreader Agreement of Prostate Imaging Reporting and Data System Version 2 Using an In-Bore MRI-Guided Prostate Biopsy Cohort: A Single Institution's Initial Experience.
Interreader Agreement of Prostate Imaging Reporting and Data System Version 2 Using an In-Bore MRI-Guided Prostate Biopsy Cohort: A Single Institution's Initial Experience.
复制标题
DOI：
10.2214/ajr.16.17551
复制
发表时间：
2017-09
期刊：
影响因子：
--
通讯作者：
Dunne RM
中科院分区：
文献类型：
--
作者：
Glazer DI;Mayo-Smith WW;Sainani NI;Sadow CA;Vangel MG;Tempany CM;Dunne RM
本研究的目的是确定前列腺成像报告和数据系统第2版（PI-RADSv 2）用于诊断前列腺癌的观察者间一致性，使用孔内MRI引导前列腺活检作为参考标准。59例患者在2010年1月21日至2013年8月21日期间接受了孔内MRI引导的前列腺活检，并在活检前6个月或更短时间内接受了诊断性多参数MRI。在回顾性审查MR图像后，为每例患者选择一处索引病变。3名接受过研究员培训的腹部放射科医生（具有1-11年经验）对临床信息设盲，根据PI-RADSv 2解释所有研究。使用Cohen kappa统计评估观察者间一致性。38处病变位于外周区，21处位于移行区。26例患者（44%）被诊断为癌症。所有活检阳性病变的总体PI-RADS评分（平均值± SD，3.9 ± 1.1）高于活检阴性病变（3.1 ± 1.0; p < 0.0001），临床显著病变（4.2 ± 1.0）高于临床不显著病变（3.1 ± 1.0; p < 0.0001）。观察者间的总体怀疑评分一致性为中度（κ = 0.45）。观察者间在外周区的总体PI-RADS评分之间存在中度一致性（κ = 0.46），在移行区存在一般一致性（κ = 0.36）。PI-RADSv 2评分在活检阳性组中较高。PI-RADSv 2显示，在没有使用评分系统经验的腹部放射科医生中，观察者之间存在中度一致性。
The purpose of this study is to determine the interobserver agreement of the Prostate Imaging Reporting and Data System version 2 (PI-RADSv2) for diagnosing prostate cancer using in-bore MRI-guided prostate biopsy as the reference standard. Fifty-nine patients underwent in-bore MRI-guided prostate biopsy between January 21, 2010, and August 21, 2013, and underwent diagnostic multiparametric MRI 6 months or less before biopsy. A single index lesion per patient was selected after retrospective review of MR images. Three fellowship-trained abdominal radiologists (with 1–11 years’ experience) blinded to clinical information interpreted all studies according to PI-RADSv2. Interobserver agreement was assessed using Cohen kappa statistics. Thirty-eight lesions were in the peripheral zone and 21 were in the transition zone. Cancer was diagnosed in 26 patients (44%). Overall PI-RADS scores were higher for all biopsy-positive lesions (mean ± SD, 3.9 ± 1.1) than for biopsy-negative lesions (3.1 ± 1.0; p < 0.0001) and for clinically significant lesions (4.2 ± 1.0) than for clinically insignificant lesions (3.1 ± 1.0; p < 0.0001). Overall suspicion score interobserver agreement was moderate (κ = 0.45). There was moderate interobserver agreement among overall PI-RADS scores in the peripheral zone (κ = 0.46) and fair agreement in the transition zone (κ = 0.36). PI-RADSv2 scores were higher in the biopsy-positive group. PI-RADSv2 showed moderate interobserver agreement among abdominal radiologists with no prior experience using the scoring system.