Molecular Imaging Branch, NCI, NIH, Bethesda, Maryland, USA (B.D.S., K.M.M., S.A.H., E.C.Y., P.L.C., B.T.); Institute of Biomedical Engineering, Department Engineering Science, University of Oxford, UK (B.D.S.).
Molecular Imaging Branch, NCI, NIH, Bethesda, Maryland, USA (B.D.S., K.M.M., S.A.H., E.C.Y., P.L.C., B.T.).
Acad Radiol. 2024 Oct;31(10):4096-4106. doi: 10.1016/j.acra.2024.04.011. Epub 2024 Apr 25.
Extraprostatic extension (EPE) is well established as a significant predictor of prostate cancer aggression and recurrence. Accurate EPE assessment prior to radical prostatectomy can impact surgical approach. We aimed to utilize a deep learning-based AI workflow for automated EPE grading from prostate T2W MRI, ADC map, and High B DWI.
An expert genitourinary radiologist conducted prospective clinical assessments of MRI scans for 634 patients and assigned risk for EPE using a grading technique. The training set and held-out independent test set consisted of 507 patients and 127 patients, respectively. Existing deep-learning AI models for prostate organ and lesion segmentation were leveraged to extract area and distance features for random forest classification models. Model performance was evaluated using balanced accuracy, ROC AUCs for each EPE grade, as well as sensitivity, specificity, and accuracy compared to EPE on histopathology.
A balanced accuracy score of .390 ± 0.078 was achieved using a lesion detection probability threshold of 0.45 and distance features. Using the test set, ROC AUCs for AI-assigned EPE grades 0-3 were 0.70, 0.65, 0.68, and 0.55 respectively. When using EPE≥ 1 as the threshold for positive EPE, the model achieved a sensitivity of 0.67, specificity of 0.73, and accuracy of 0.72 compared to radiologist sensitivity of 0.81, specificity of 0.62, and accuracy of 0.66 using histopathology as the ground truth.
Our AI workflow for assigning imaging-based EPE grades achieves an accuracy for predicting histologic EPE approaching that of physicians. This automated workflow has the potential to enhance physician decision-making for assessing the risk of EPE in patients undergoing treatment for prostate cancer due to its consistency and automation.
前列腺外延伸（EPE）是预测前列腺癌侵袭性和复发的重要指标。在根治性前列腺切除术前准确评估 EPE 可影响手术方式。我们旨在利用基于深度学习的 AI 工作流程，从前列腺 T2W MRI、ADC 图和高 b 值 DWI 自动分级 EPE。
一名泌尿生殖系统放射科专家对 634 例患者的 MRI 扫描进行了前瞻性临床评估，并使用分级技术评估了 EPE 的风险。训练集和独立测试集由 507 例和 127 例患者组成。利用现有的前列腺器官和病变分割深度学习 AI 模型，提取用于随机森林分类模型的面积和距离特征。使用平衡准确率、每个 EPE 分级的 ROC AUC 以及与组织病理学比较的 EPE 的敏感性、特异性和准确性来评估模型性能。
使用病变检测概率阈值为 0.45 和距离特征，得到平衡准确率为 0.390±0.078。使用测试集，AI 分配的 EPE 分级 0-3 的 ROC AUC 分别为 0.70、0.65、0.68 和 0.55。当使用 EPE≥1 作为阳性 EPE 的阈值时，与病理学家使用组织病理学作为金标准的 0.81 的敏感性、0.62 的特异性和 0.66 的准确性相比，该模型的敏感性为 0.67、特异性为 0.73 和准确性为 0.72。
我们用于分配基于影像学的 EPE 分级的 AI 工作流程，在预测组织学 EPE 方面的准确性接近医生。由于其一致性和自动化，这种自动工作流程有可能增强医生评估接受前列腺癌治疗患者 EPE 风险的决策能力。
