• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Society of Photo-Optical Instrumentation Engineers;SPIE Medical Imaging Conference >Automatic high-grade cancer detection on prostatectomy histopathology images
【24h】

Automatic high-grade cancer detection on prostatectomy histopathology images

机译：前列腺切除术组织病理学图像上的自动高级癌症检测

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Automatic cancer grading and high-grade cancer detection for radical prostatectomy (RP) specimens can benefitpathological assessment for prognosis and post-surgery treatment decision making. We developed and validated anautomatic system which grades cancerous tissue as high-grade (Gleason grade 4 and higher) vs. low-grade (Gleasongrade 3) on digital histopathology whole-slide images (WSIs). We combined this grading system with our previouslyreportedcancer detection system to build a high-grade cancer detection system which automatically finds high-gradecancerous foci on WSIs. The system was tuned on a 3-patient data set and cross-validated against expert-drawn contourson a separate 68-patient data set comprising 286 mid-gland whole-slide images of RP specimens. The system usesmachine learning techniques to classify each region of interest (ROI) on the slide as cancer or non-cancer and eachcancerous ROI as high-grade or low-grade cancer. We used leave-one-patient-out cross-validation to measure theperformance of cancer grading for classified ROIs with three different classifiers and the performance of the high-gradecancer detection system on a per tumor focus basis. The best performing (Fisher) classifier yielded an area under thereceiver-operating characteristic curve of 0.87 for cancer grading. The system yielded error rates of 19.5% and 23.4% forpure high-grade (Gleason 4+4, 5+5) and high-grade (Gleason Score ≥ 7) cancer detection, respectively. The systemdemonstrated potential for practical computation speeds. Upon successful multi-centre validation, this system has thepotential to assist the pathologist to find high-grade cancer more efficiently, which benefits the selection and guidance ofadjuvant therapy and prognosis post RP.
机译：根治性前列腺切除术（RP）标本的自动癌症分级和高级癌症检测可以受益 病理评估以进行预后和手术后治疗决策。我们开发并验证了 自动系统，将癌组织分级为低等级（格里森4级及更高），将其分为低等级（格里森） 3年级）上的数字组织病理学全幻灯片图像（WSI）。我们将此评分系统与我们先前报告的相结合 癌症检测系统，以构建自动发现高品位的高级癌症检测系统 WSI的癌灶。该系统在3位患者的数据集上进行了调试，并针对专家绘制的轮廓进行了交叉验证 在包含286个RP样本的中腺全幻灯片图像的单独的68位患者数据集上。系统使用 机器学习技术，将幻灯片上的每个感兴趣区域（ROI）归类为癌症或非癌症，并且每个 ROI为高级别或低级别癌症。我们使用了“请一位病人离开”交叉验证来衡量 三种不同分类器的已分类ROI的癌症分级性能和高级别的性能 基于每个肿瘤焦点的癌症检测系统。效果最好的（Fisher）分类器在 癌症分级的受试者工作特性曲线为0.87。该系统的错误率分别为19.5％和23.4％ 纯高等级（格里森4 + 4、5 + 5）和高等级（格里森评分≥7）癌症检测。系统 展示了实用计算速度的潜力。在成功进行多中心验证后，该系统将 协助病理学家更有效地发现高级别癌症的潜力，这有利于选择和指导 RP后的辅助治疗和预后。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号