文献速递前列腺癌

深度学习用于前列腺癌根治术中荧光共聚焦显微镜图像判读

Deep learning for fluorescence confocal microscopy image interpretation in radical prostatectomy.
PubMed 文献 · BJU Int · 2026-04-19

本研究基于多中心IP8-FLUORESCE研究数据,开发并验证了用于根治性前列腺切除术中荧光共聚焦显微镜(FCM)图像手术切缘评估的深度学习模型。内部测试集(n=57)灵敏度87.5%、特异度97.9%、AUC 0.93;外部验证集(46张独立图像)灵敏度91.3%、特异度73.9%、AUC 0.83。模型可通过图形界面在2秒内给出可解释预测,有望减少对术中病理支持的依赖。

要点

  • 内部测试AUC 0.93,外部验证AUC 0.83,显示较强判别力与泛化性。
  • 外部验证灵敏度91.3%但特异度降至73.9%,需警惕假阳性。
  • Grad-CAM热图与恶性结构对齐,GUI支持2秒内实时判读。

核心结论

  • 深度学习模型对FCM图像判别性能强且具泛化性,可作为术中切缘评估的可扩展方案。

研究设计与样本

  • 基于IP8-FLUORESCE多中心研究FCM图像训练和测试卷积神经网络,以常规组织病理切缘状态为参考标准。
  • 模型开发与内部测试共纳入24例患者的275张图像(37张肿瘤、238张良性)。
  • 外部验证使用LaserSAFE可行性试验的46张独立图像。

主要发现

  • 内部测试集(n=57)灵敏度87.5%、特异度97.9%、AUC 0.93,Brier评分0.16。
  • 外部验证灵敏度91.3%、特异度73.9%、AUC 0.83,Brier评分0.20。
  • Grad-CAM可视化与FCM图像中恶性结构对齐,GUI预测时间小于2秒。

关键数值与推荐

  • 图像预处理提取特定数字缩放级别感兴趣区并归一化为896×896像素。
  • 模型改进采用焦点损失结合标签平滑、dropout正则化、自适应类别加权和加权采样以应对类别不平衡。

临床怎么用

  • 该模型可用于根治性前列腺切除术中实时FCM图像判读,辅助手术切缘评估。
  • GUI支持快速可解释预测,可能减少对术中病理支持的依赖。

局限性

  • 外部验证特异度73.9%低于内部测试,假阳性风险需进一步评估。
  • 模型开发与内部测试样本量有限(24例患者、275张图像)。

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

OBJECTIVE: To develop and validate a deep learning model for interpretation of fluorescence confocal microscopy (FCM) images for intraoperative surgical margin assessment during radical prostatectomy (RP). PATIENTS AND METHODS: Fluorescence confocal microscopy images from the multicentre Imperial Prostate 8-Fluorescence Confocal Microscopy for Rapid Evaluation of Surgical Cancer Excision (IP8-FLUORESCE) study were used to train and test a convolutional neural network model. The modified model incorporated focal loss with label smoothing, dropout regularisation, adaptive class weighting, and weighted sampling to address pronounced class imbalance. Images were pre-processed by extracting regions of interest at a defined digital zoom level and normalised to 896 × 896 pixels. The reference standard was surgical margin status on conventional histopathology assessed by an expert histopathologist. Diagnostic performance was assessed using sensitivity, specificity, positive and negative predictive value, area under the receiver-operating-characteristic curve (AUC), and calibration via Brier scores. External validation was conducted using an independent dataset from the LaserSAFE feasibility trial. Model explainability was evaluated using Gradient-weighted Class Activation Mapping (Grad-CAM) and a custom graphical user interface (GUI) was developed to support real-time deployment. RESULTS: A total of 275 images (37 tumour and 238 benign from 24 patients) were included for model development and internal testing. On the internal test set (n = 57), the model achieved a sensitivity of 87.5%, specificity of 97.9%, and an AUC of 0.93, with good calibration (Brier score 0.16). External validation using 46 independent images yielded a sensitivity of 91.3%, specificity of 73.9%, and an AUC of 0.83, with acceptable calibration (Brier score 0.20). Grad-CAM visualisations aligned with malignant structures on FCM images, and the GUI enabled rapid, interpretable predictions in <2 s. CONCLUSIONS: We developed and validated a deep learning model for interpretation of FCM images from RP specimens, which demonstrated strong discriminative performance and generalisability for automated FCM interpretation. This approach represents a scalable solution for real-time intraoperative margin assessment and may reduce reliance on intraoperative pathology support.

来源:BJU Int,查看原文。版权归原作者所有。

引用

  1. Fang L, Mayor N, Light A, et al. Deep learning for fluorescence confocal microscopy image interpretation in radical prostatectomy. BJU Int. 2026. PMID: 42001901. doi:10.1111/bju.70273

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。