Department of Radiology, Abdominal Imaging Division, Massachusetts General Hospital, 55 Fruit Street, White 270, Boston, MA 02114; Harvard Medical School, 55 Fruit Street, White 270, Boston, MA 02114.
Department of Radiology, Institute for Technology Assessment, Massachusetts General Hospital, Boston, Massachusetts (S.M.).
Acad Radiol. 2023 Jul;30(7):1340-1349. doi: 10.1016/j.acra.2022.09.009. Epub 2022 Oct 7.
To evaluate whether addition of a computer-aided diagnostic (CAD) generated MRI series improves detection of clinically significant prostate cancer.
Nine radiologists retrospectively interpreted 150 prostate MRI examinations without and then with an additional random forest-based CAD model-generated MRI series. Characteristics of biopsy negative versus positive (Gleason ≥ 7 adenocarcinoma) groups were compared using the Wilcoxon test for continuous and Pearson's chi-squared test for categorical variables. The diagnostic performance of readers was compared without versus with CAD using MRMC methods to estimate the area under the receiver operator characteristic curve (AUC). Inter-reader agreement was assessed using weighted inter-rater agreement statistics. Analyses were repeated in peripheral and transition zone subgroups.
Among 150 men with median age 67 ± 7.4 years, those with clinically significant prostate cancer were older (68 ± 7.6 years vs. 66 ± 7.0 years; p < .02), had smaller prostate volume (43.9 mL vs. 60.6 mL; p < .001), and no difference in prostate specific antigen (PSA) levels (7.8 ng/mL vs. 6.9 ng/mL; p = .08), but higher PSA density (0.17 ng/mL/cc vs. 0.10 ng/mL/cc; p < .001). Inter-rater agreement (IRA) for PI-RADS scores was moderate without CAD and significantly improved to substantial with CAD (IRA = 0.47 vs. 0.65; p < .001). CAD also significantly improved average reader AUC (AUC = 0.72, vs. AUC = 0.67; p = .02).
Addition of a random forest method-based, CAD-generated MRI image series improved inter-reader agreement and diagnostic performance for detection of clinically significant prostate cancer, particularly in the transition zone.
评估计算机辅助诊断（CAD）生成的 MRI 序列是否能提高临床显著前列腺癌的检出率。
9 位放射科医生回顾性分析了 150 例前列腺 MRI 检查，首先不使用，然后使用基于随机森林的 CAD 模型生成的 MRI 序列。使用 Wilcoxon 检验对连续变量和 Pearson's chi-squared 检验对分类变量进行比较，比较活检阴性和阳性（Gleason ≥ 7 级腺癌）组之间的特征。使用 MRMC 方法比较有无 CAD 时读者的诊断性能，以估计受试者工作特征曲线下的面积（AUC）。使用加权组内一致性统计量评估读者间的一致性。在周边区和移行区亚组中重复分析。
在 150 名中位年龄为 67 ± 7.4 岁的男性中，患有临床显著前列腺癌的患者年龄更大（68 ± 7.6 岁比 66 ± 7.0 岁；p <.02），前列腺体积更小（43.9 mL 比 60.6 mL；p <.001），前列腺特异性抗原（PSA）水平无差异（7.8 ng/mL 比 6.9 ng/mL；p =.08），但 PSA 密度更高（0.17 ng/mL/cc 比 0.10 ng/mL/cc；p <.001）。无 CAD 时 PI-RADS 评分的读者间一致性（IRA）为中等，而 CAD 时显著提高到较高（IRA = 0.47 比 0.65；p <.001）。CAD 还显著提高了平均读者 AUC（AUC = 0.72，AUC = 0.67；p =.02）。
添加基于随机森林方法的 CAD 生成的 MRI 图像序列提高了读者间的一致性和诊断性能，可提高临床显著前列腺癌的检出率，尤其是在移行区。
