基于扩散基础谱成像指标的人工智能模型可准确预测临床显著性前列腺癌
研究纳入2020年2月至2024年3月活检前接受含常规与DBSI序列前列腺MRI的241例患者,以活检病理为金标准构建DBSI指标人工智能模型预测临床显著性前列腺癌(Gleason评分>7)。DBSI模型为独立预测因子(OR 2.04,95%CI 1.52-2.73,P<.01),单独效能与PSAD+PI-RADS相似(AUC 0.863 vs 0.859,P=.89),联合PI-RADS判别力最高(AUC 0.894,P<.01);对PI-RADS 1-3患者采用该模型策略可减少27%活检并漏诊2% csPCa。
要点
- DBSI人工智能模型是csPCa的独立预测因子(OR 2.04,95%CI 1.52-2.73,P<.01)。
- DBSI模型单独判别力与PSAD+PI-RADS相似(AUC 0.863 vs 0.859,P=.89)。
- DBSI模型+PI-RADS判别力最高(AUC 0.894,P<.01),PI-RADS 1-3者用该策略可减少27%活检、漏诊2% csPCa。
核心结论
- 基于DBSI的人工智能模型能准确预测活检证实的csPCa,并可联合PI-RADS有望减少不必要的前列腺穿刺。
研究设计与样本
- 2020年2月至2024年3月,241例患者在 prostate biopsy 前接受包含常规与DBSI特定序列的前列腺MRI。
- 以DBSI指标为输入构建人工智能分类模型,以活检病理为金标准,并与PSA、PSAD、PI-RADS比较风险判别能力。
- csPCa定义为Gleason评分>7。
主要发现
- 校正年龄、家族史和种族后,DBSI模型为csPCa独立预测因子(OR 2.04,95%CI 1.52-2.73,P<.01)。
- PSAD(OR 2.02,95%CI 1.21-3.35,P=.01)与PI-RADS分类亦为独立预测因子:PI-RADS 3的OR 4.00(95%CI 1.37-11.6,P=.01),PI-RADS 4-5的OR 9.67(95%CI 2.89-32.7,P<.01)。
- 在本数据集内,DBSI模型单独与PSAD+PI-RADS判别力相似(AUC 0.863 vs 0.859,P=.89)。
- DBSI模型+PI-RADS联合的csPCa风险判别力最高(AUC 0.894,P<.01)。
- 对PI-RADS 1-3患者采用DBSI模型策略,与全部活检相比可减少27%活检,同时漏诊2%的csPCa。
临床怎么用
- 作者结论认为DBSI人工智能模型可准确预测活检csPCa,并可联合PI-RADS有望减少不必要的前列腺穿刺。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
PURPOSE: Conventional prostate magnetic resonance imaging has limited accuracy for clinically significant prostate cancer (csPCa). We performed diffusion basis spectrum imaging (DBSI) before biopsy and applied artificial intelligence models to these DBSI metrics to predict csPCa. MATERIALS AND METHODS: Between February 2020 and March 2024, 241 patients underwent prostate MRI that included conventional and DBSI-specific sequences before prostate biopsy. We used artificial intelligence models with DBSI metrics as input classifiers and the biopsy pathology as the ground truth. The DBSI-based model was compared with available biomarkers (PSA, PSA density [PSAD], and Prostate Imaging Reporting and Data System [PI-RADS]) for risk discrimination of csPCa defined as Gleason score > 7. RESULTS: The DBSI-based model was an independent predictor of csPCa (odds ratio [OR] 2.04, 95% CI 1.52-2.73, P < .01), as were PSAD (OR 2.02, 95% CI 1.21-3.35, P = .01) and PI-RADS classification (OR 4.00, 95% CI 1.37-11.6 for PI-RADS 3, P = .01; OR 9.67, 95% CI 2.89-32.7 for PI-RADS 4-5, P < .01), adjusting for age, family history, and race. Within our dataset, the DBSI-based model alone performed similarly to PSAD + PI-RADS (AUC 0.863 vs 0.859, P = .89), while the combination of the DBSI-based model + PI-RADS had the highest risk discrimination for csPCa (AUC 0.894, P < .01). A clinical strategy using the DBSI-based model for patients with PI-RADS 1-3 could have reduced biopsies by 27% while missing 2% of csPCa (compared with biopsy for all). CONCLUSIONS: Our DBSI-based artificial intelligence model accurately predicted csPCa on biopsy and can be combined with PI-RADS to potentially reduce unnecessary prostate biopsies.
引用
- Kim EH, Jing H, Utt KL, et al. An Artificial Intelligence Model Using Diffusion Basis Spectrum Imaging Metrics Accurately Predicts Clinically Significant Prostate Cancer. J Urol. 2025. PMID: 39869819. doi:10.1097/JU.0000000000004456
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。