卵巢癌转移灶的智能识别与结构化报告填充：一项多中心研究

Intelligent identification of ovarian cancer metastases and a structured population report: a multicenter study

摘要

目的 探索人工智能技术在卵巢癌转移灶精准定位和评估中的应用模式。方法 共纳入3个中心273例卵巢癌转移患者腹盆腔增强计算机体层成像（CT）图像，经诊断医师标注，共获得174个膈下转移灶及516个肝周转移灶，随机划分为训练集（n=561）和测试集（n=129），构建基于深度卷积网络的膈下/肝周位置二分类模型，计算其准确率、灵敏度、特异度、精确度、 F1值及曲线下面积（AUC）。基于增强CT四期图像及手术病理资料，填充结构化报告并评估其性能。结果 膈下/肝周位置区分模型的AUC为0.78，准确率0.721，灵敏度0.417，特异度0.839，精确度0.500， F1值0.455。结构化报告填充中，对肝周转移灶位置的分类模型表现最佳，AUC为0.83，准确率0.753，灵敏度0.804，特异度0.702，精确度0.725， F1值0.763；其他特征模型的识别能力有待提升。结论 本研究探索并构建了“影像自动分析-关键特征提取-报告结构化填充”的临床辅助工作模式，为优化诊断流程、提升报告标准化水平提供了实践框架。

Abstract

Objective To explore application models of artificial intelligence for precise localization and evaluation of ovarian cancer metastases. Methods A total of 273 contrast-enhanced abdominal-pelvic computed tomography (CT) scans from patients with ovarian cancer metastases across three centers were included. Radiologists annotated 174 subdiaphragmatic metastases and 516 perihepatic metastases, who were randomly divided into training (n=561) and test (n=129) sets. A deep convolutional network-based binary classification model for distinguishing subdiaphragmatic/perihepatic locations was constructed, and its accuracy, sensitivity, specificity, precision, F1-score, and area under the curve (AUC) were calculated. Using four-phase contrast-enhanced CT images and surgical pathology data, the structured reports were filled in and their performance evaluated. Results The subphrenic/perihepatic location differentiation model achieved an AUC of 0.78, with an accuracy of 0.721, sensitivity of 0.417, specificity of 0.839, precision of 0.500, and an F1-score of 0.455. In the structured report population, the classification model for perihepatic metastasis location performed best, attaining an AUC of 0.83, accuracy of 0.753, sensitivity of 0.804, specificity of 0.702, precision of 0.725, and an F1-score of 0.763. The recognition capabilities of models for other features require further improvement. Conclusion This work establishes a novel clinical assistance workflow—"automated image analysis, key feature extraction, and structured report population"—offering a practical framework for optimizing diagnostic processes and enhancing reporting standardization.

关键词

卵巢癌 / 腹膜转移 / 计算机体层成像 / 人工智能 / 结构化报告

Key words

ovarian cancer / peritoneal metastasis / computed tomography / artificial intelligence / structured reporting

参考文献

[1] ARMSTRONG D K, ALVAREZ R D, BACKES F J, et al. Nccn guidelines(R) insights: ovarian cancer, version 3.2022 featured updates to the NCCN guidelines[J]. Journal of the national comprehensive cancer network, 2022, 20(9): 972-980.

[2] SIEGEL R L, GIAQUINTO A N, JEMAL A. Cancer statistics, 2024[J]. Ca: a cancer journal for clinicians, 2024, 74(1): 12-49.

[3] YOUSEFI M, DEHGHANI S, NOSRATI R, et al. Current insights into the metastasis of epithelial ovarian cancer—hopes and hurdles[J]. Cellular oncology (dordrecht, Netherlands), 2020, 43(4): 515-538.

[4] LEDERMANN J A, MATIAS—GUIU X, AMANT F, et al. Esgo—esmo—esp consensus conference recommendations on ovarian cancer: pathology and molecular biology and early, advanced and recurrent disease[J]. Annals of oncology: official journal of the European Society for Medical Oncology, 2024, 35(3): 248-266.

[5] JEONG S Y, KIM T J, PARK B K. Epithelial ovarian cancer: a review of preoperative imaging features indicating suboptimal surgery[J]. Journal of gynecologic oncology, 2020, 31(4): e57.

[6] DE OLIVEIRA L R L B, HORVAT N, ANDRIEU P I C, et al. Ovarian cancer staging: what the surgeon needs to know[J]. The British journal of radiology, 2021, 94(1125): 20210091.

[7] 秦萌, 金滢, 潘凌亚. 晚期卵巢癌患者新辅助化疗适应证及间歇性肿瘤细胞减灭术手术时机选择[J]. 中国实用妇科与产科杂志, 2018, 34(2): 232-236.

[8] SUIDAN R S, RAMIREZ P T, SARASOHN D M, et al. A multicenter assessment of the ability of preoperative computed tomography scan and CA—125 to predict gross residual disease at primary debulking for advanced epithelial ovarian cancer[J]. Gynecologic oncology, 2017, 145(1): 27-31.

[9] KANG S K, REINHOLD C, ATRI M , ET AL. Acr appropriateness criteria(R) staging and follow—up of ovarian cancer[J]. Journal of the American college of radiology, 2018, 15(5): S198-S207.

[10] O'NEILL A C, SOMAROUTHU B, TIRUMANI S H, et al. Patterns and prognostic importance of hepatic involvement in patients with serous ovarian cancer: a single—institution experience with 244 patients[J]. Radiology, 2017, 282(1): 160-170.

[11] AKIN O, SALA E, MOSKOWITZ C S, et al. Perihepatic metastases from ovarian cancer: sensitivity and specificity of CT for the detection of metastases with and those without liver parenchymal invasion[J]. Radiology, 2008, 248(2): 511-517.

[12] 王坚. 从外科角度透视卵巢癌肿瘤细胞减灭术的关键环节[J]. 上海医学, 2023, 46(11): 730-734.

[13] LU J J, CAI S Q, WANG F, et al. Development of a prediction model for gross residual in high—grade serous ovarian cancer by combining preoperative assessments of abdominal and pelvic metastases and multiparametric MRI[J]. Academic radiology, 2023, 30(9): 1823-1831.

[14] 李海明, 郭勤浩, 陆静, 等. 晚期卵巢癌术前影像学评估研究进展[J]. 肿瘤影像学, 2025, 34(4): 301-311.

[15] AI Y, ZHANG J D, JIN J B, et al. Preoperative prediction of metastasis for ovarian cancer based on computed tomography radiomics features and clinical factors[J]. Frontiers in oncology, 2021, 11: 610742.

[16] ANDRIEU P C, NIKOLOVSKI I, JULURU K, et al. Synoptic reporting for pretreatment ct examination in patients with advanced ovarian cancer: impact on documentation of disease sites and physician satisfaction[J]. American journal of roentgenology, 2023, 221(6): 760-772.

[17] HEWER E. The oncologist’s guide to synoptic reporting: a primer[J]. Oncology, 2020, 98(6): 396-402.

[18] YAO W J, BAI J J, LIAO W, et al. From CNN to transformer: a review of medical image segmentation models[J]. Journal of imaging informatics in medicine, 2024, 37(4): 1529-1547.

[19] BUDDENKOTTE T, RUNDO L, WOITEK R, et al. Deep learning—based segmentation of multisite disease in ovarian cancer[J]. European radiology experimental, 2023, 7(1): 77.

[20] KARPATHY A, LI F F. Deep visual—semantic alignments for generating image descriptions[J]. IEEE transactions on pattern analysis and machine intelligence, 2017, 39(4): 664-676.

[21] HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]// Proceedings of the IEEE conference on computer vision and pattern recognition, Las Vegas: IEEE, 2016: 770-778.

[22] 张国楠, 郎景和. 卵巢癌腹腔镜手术所致腹壁穿刺孔种植转移的肿瘤学与医学伦理思考[J]. 肿瘤预防与治疗, 2025, 38(4): 263-266.

[23] 雷婷, 曹梦祯, 李君芬, 等. 深度学习在子宫内膜癌影像诊断中的研究进展[J]. 兰州大学学报（医学版）, 2025, 51(11): 81-87.

基金资助

中国医学科学院医学与健康科技创新工程项目(2024-I2M-C&T-B-032)

中央高水平医院临床科研业务费(2025-PUMCH-A-023)