深度学习超声影像组学列线图模型鉴别Ⅰ和Ⅱ型上皮性卵巢癌

doi:10.3969/j.issn.1006-5725.2025.18.020

Abstract

Abstract:

Objective To evaluate an ultrasound-based deep learning radiomics nomogram （DLR_Nomogram） for non-invasively differentiating between type Ⅰ and type Ⅱ epithelial ovarian cancer （EOC） before surgery. Methods In this study， a cohort of 195 patients diagnosed with EOC was analyzed. Participants were randomly divided into a training set and a testing set at an 8∶2 ratio. Following data preprocessing， region of interest （ROI） delineation， feature extraction and selection， as well as the clipping and extraction of the maximum section sonogram for each sample， three initial models were developed： the radiomics signature （Rad_Sig）， the deep transfer learning signature （DTL_Sig）， and the clinical signature （Clinic_Sig）. Subsequently， an integrated model—referred to as the DLR_Nomogram—was constructed by combining Rad_Sig， DTL_Sig， and Clinic_Sig， and was presented in the form of a nomogram. The performance of the model was evaluated using the receiver operating characteristic （ROC） curve and the corresponding area under the curve （AUC）. Results In the testing set， the DLR_Nomogram demonstrated superior predictive performance （AUC： 0.951， 95%CI： 0.876 ~ 1.000） compared to Rad_Sig （AUC： 0.709， 95%CI： 0.539 ~ 0.880）， DTL_Sig （AUC： 0.842， 95%CI： 0.712 ~ 0.972）， and Clinic_Sig （AUC： 0.916， 95%CI： 0.827 ~ 1.000）. The Hosmer?Lemeshow goodness-of-fit test for the DLR_Nomogram resulted in a p-value exceeding 0.05， indicating adequate model calibration. Moreover， decision curve analysis revealed that the DLR_Nomogram offers a higher net clinical benefit across a defined range of threshold probabilities. Conclusions The ultrasound-based DLR_Nomogram exhibits a robust ability to differentiate between Type Ⅰ and Type Ⅱ EOC， and may serve as a valuable clinical tool for guiding individualized preoperative diagnostic and therapeutic decision-making.

Key words: epithelial ovarian cancer, ultrasound, radiomics, deep transfer learning, nomogram

CLC Number:

R711.75

Yangchun DU,Hongyu ZHENG,Haining CHEN,Wenwen GUO,Jinxiu YAO,Tongliu LAN,Yanju XIAO. Ultrasound⁃based deep learning radiomics nomogram to differentiate type Ⅰ and type Ⅱ epithelial ovarian cancer[J]. The Journal of Practical Medicine, 2025, 41(18): 2920-2927.

Figures/Tables 8

Fig.1

Tab.1

Clinical parameters and sonographic semantic features of the training and testing sets"

项目	训练集（n = 156）				测试集（n = 39）
项目	Ⅰ型EOC（n = 64）	Ⅱ型EOC（n = 92）	t/χ² 值	P值	Ⅰ型EOC （n = 16）	Ⅱ型EOC （n = 23）	t/χ² 值	P值
年龄（x ± s）/岁	47.31 ± 10.34	54.04 ± 9.66	-4.158	< 0.001	50.06 ± 10.32	56.17 ± 8.10	-2.07	0.045
单核细胞计数（x ± s）/（10⁹·L^-1）	0.60 ± 0.27	0.59 ± 0.27	0.228	0.747	0.45 ± 0.19	0.73 ± 0.24	-3.89	< 0.001
血小板计数（x ± s）/（10⁹·L^-1）	316.03 ± 88.64	365.91 ± 110.79	-2.995	0.004	298.19 ± 75.92	368.17 ± 101.86	-2.331	0.025
PLR（x ± s）	200.27 ± 128.317	242.95 ± 130.71	-2.021	0.045	199.09 ± 143.20	252.62 ± 118.34	-1.275	0.21
LMR（x ± s）	3.49 ± 1.63	3.66 ± 2.88	-0.427	0.662	4.37 ± 1.93	2.46 ± 1.08	3.952	< 0.001
FIGO分期			30.628	< 0.001			11.493	< 0.001
Ⅰ+Ⅱ	49（76.56）	29（31.52）			13（81.25）	6（26.09）
Ⅲ+Ⅳ	15（23.44）	63（68.48）			3（18.75）	17（73.91）
月经状态			9.748	0.002			0.321	0.583
绝经前	42（65.62）	37（40.22）			7（43.75）	8（34.78）
绝经后	22（34.38）	55（59.78）			9（56.25）	15（65.22）
肿块声像图特征			21.361	< 0.001			5.252	0.020
囊性	10（15.62）	1（1.09）			2（12.50）	0（0.00）
囊实性	38（59.38）	40（43.48）			10（62.50）	10（43.48）
实性	16（25.00）	51（55.43）			4（25.00）	13（56.52）
彩色血流评分			17.311	< 0.001			4.503	0.032
1分	6（9.38）	3（3.26）			1（6.25）	0（0.00）
2分	38（59.38）	30（32.61）			8（50.00）	6（26.09）
3分	19（29.69）	52（56.52）			7（43.75）	16（69.57）
4分	1（1.56）	7（7.61）			0（0.00）	1（4.35）
肿瘤侧别			17.11	< 0.001			5.564	0.018
单侧	51（79.69）	43（46.74）			13（81.25）	10（43.48）
双侧	13（20.31）	49（53.26）			3（18.75）	13（56.52）
CA125			18.686	< 0.001			8.474	0.002
≤ 35 U/mL	19（29.69）	7（7.61）			8（50.00）	2（8.70）
> 35 U/mL且≤ 200 U/mL	21（32.81）	22（23.91）			3（18.75）	4（17.39）
> 200 U/mL且> 500 U/mL	8（12.50）	19（20.65）			2（12.50）	5（21.74）
≥ 500 U/mL	16（25.00）	44（47.83）			3（18.75）	12（52.17）

Tab.1

Fig.2

Fig.3

Tab.2

Fig.4

Fig.5

Fig.6

References 27

[1]	张丹，李燕东. 卵巢与附件区肿瘤的研究进展［J］. 中华医学超声杂志（电子版）， 2020， 17（3）： 274-278.
[2]	QIAN L， REN J， LIU A， et al. MR imaging of epithelial ovarian cancer： A combined model to predict histologic subtypes［J］. Eur Radiol， 2020， 30（11）： 5815-5825. doi:10.1007/s00330-020-06993-5 doi: 10.1007/s00330-020-06993-5
[3]	BERA K， BRAMAN N， GUPTA A， et al. Predicting cancer outcomes with radiomics and artificial intelligence in radiology［J］. Nat Rev Clin Oncol， 2022， 19（2）： 132-146. doi:10.1038/s41571-021-00560-7 doi: 10.1038/s41571-021-00560-7
[4]	LI X， YANG L， JIAO X. Comparison of Traditional Radiomics， Deep Learning Radiomics and Fusion Methods for Axillary Lymph Node Metastasis Prediction in Breast Cancer［J］. Acad Radiol， 2023， 30（7）： 1281-1287. doi:10.1016/j.acra.2022.10.015 doi: 10.1016/j.acra.2022.10.015
[5]	PARK H J， PARK B， LEE S S. Radiomics and Deep Learning： Hepatic Applications［J］. Korean J Radiol， 2020， 21（4）： 387-401. doi:10.3348/kjr.2019.0752 doi: 10.3348/kjr.2019.0752
[6]	罗刚，泮思林，乔思波，等. 深度学习技术在胎儿超声心动图图像自动识别中的应用［J］. 实用医学杂志， 2022， 38（14）： 1830-1833. doi:10.3969/j.issn.1006⁃5725.2022.14.022 doi: 10.3969/j.issn.1006?5725.2022.14.022
[7]	TONG T， GU J， XU D， et al. Deep learning radiomics based on contrast-enhanced ultrasound images for assisted diagnosis of pancreatic ductal adenocarcinoma and chronic pancreatitis［J］. BMC Med， 2022， 20（1）： 74. doi:10.1186/s12916-022-02258-8 doi: 10.1186/s12916-022-02258-8
[8]	SOHN J H， FIELDS B K K. Radiomics and Deep Learning to Predict Pulmonary Nodule Metastasis at CT［J］. Radiology， 2024， 311（1）： e233356. doi:10.1148/radiol.233356 doi: 10.1148/radiol.233356
[9]	DU Y， XIAO Y， GUO W， et al. Development and validation of an ultrasound-based deep learning radiomics nomogram for predicting the malignant risk of ovarian tumours［J］. Biomed Eng Online， 2024， 23（1）： 41. doi:10.1186/s12938-024-01234-y doi: 10.1186/s12938-024-01234-y
[10]	DU Y， GUO W， XIAO Y， et al. Ultrasound-based deep learning radiomics model for differentiating benign， borderline， and malignant ovarian tumours： A multi-class classification exploratory study［J］. BMC Med Imaging， 2024， 24（1）： 89. doi:10.1186/s12880-024-01251-2 doi: 10.1186/s12880-024-01251-2
[11]	CHEN H， YANG B W， QIAN L， et al. Deep Learning Prediction of Ovarian Malignancy at US Compared with O-RADS and Expert Assessment［J］. Radiology， 2022， 304（1）： 106-113. doi:10.1148/radiol.211367 doi: 10.1148/radiol.211367
[12]	JIAN J， LI Y， PICKHARDT P J， et al. MR image-based radiomics to differentiate type Ⅰ and type Ⅱ epithelial ovarian cancers［J］. Eur Radiol， 2021， 31（1）： 403-410. doi:10.1007/s00330-020-07091-2 doi: 10.1007/s00330-020-07091-2
[13]	ZHANG H， MAO Y， CHEN X， et al. Magnetic resonance imaging radiomics in categorizing ovarian masses and predicting clinical outcome： A preliminary study［J］. Eur Radiol， 2019， 29（7）： 3358-3371. doi:10.1007/s00330-019-06124-9 doi: 10.1007/s00330-019-06124-9
[14]	CHEN X， WANG X， ZHANG K， et al. Recent advances and clinical applications of deep learning in medical image analysis［J］. Med Image Anal， 2022， 79： 102444. doi:10.1016/j.media.2022.102444 doi: 10.1016/j.media.2022.102444
[15]	郭润财，王蕾，黄振国，等. 基于从头训练模式深度学习卷积神经网络模型评估急性肺栓塞的价值［J］. 实用医学杂志， 2023， 39（22）： 2979-2983. doi:10.3969/j.issn.1006-5725.2023.22.021 doi: 10.3969/j.issn.1006-5725.2023.22.021
[16]	TOSEEF M， OLAYEMI PETINRIN O， WANG F， et al. Deep transfer learning for clinical decision-making based on high-throughput data： Comprehensive survey with benchmark results［J］. Brief Bioinform， 2023， 24（4）： 1-16. doi:10.1093/bib/bbad254 doi: 10.1093/bib/bbad254
[17]	WANG R， CAI Y， LEE I K， et al. Evaluation of a convolutional neural network for ovarian tumor differentiation based on magnetic resonance imaging［J］. Eur Radiol， 2021， 31（7）： 4960-4971. doi:10.1007/s00330-020-07266-x doi: 10.1007/s00330-020-07266-x
[18]	CHRISTIANSEN F， EPSTEIN E L， SMEDBERG E， et al. Ultrasound image analysis using deep neural networks for discriminating between benign and malignant ovarian tumors： Comparison with expert subjective assessment［J］. Ultrasound Obstet Gynecol， 2021， 57（1）： 155-163. doi:10.1002/uog.23530 doi: 10.1002/uog.23530
[19]	GAO Y， ZENG S， XU X， et al. Deep learning-enabled pelvic ultrasound images for accurate diagnosis of ovarian cancer in China： A retrospective， multicentre， diagnostic study［J］. Lancet Digit Health， 2022， 4（3）： e179-e187. doi:10.1016/s2589-7500(21)00278-8 doi: 10.1016/s2589-7500(21)00278-8
[20]	HSU S T， SU Y J， HUNG C H， et al. Automatic ovarian tumors recognition system based on ensemble convolutional neural network with ultrasound imaging［J］. BMC Med Inform Decis Mak， 2022， 22（1）： 298. doi:10.1186/s12911-022-02047-6 doi: 10.1186/s12911-022-02047-6
[21]	WANG T， WANG H， WANG Y， et al. MR-based radiomics-clinical nomogram in epithelial ovarian tumor prognosis prediction： Tumor body texture analysis across various acquisition protocols［J］. J Ovarian Res， 2022， 15（1）： 1-10. doi:10.1186/s13048-021-00941-7 doi: 10.1186/s13048-021-00941-7
[22]	ZHU H， AI Y， ZHANG J， et al. Preoperative Nomogram for Differentiation of Histological Subtypes in Ovarian Cancer Based on Computer Tomography Radiomics［J］. Front Oncol， 2021， 11： 642892. doi:10.3389/fonc.2021.642892 doi: 10.3389/fonc.2021.642892
[23]	YAN B C， MA X L， LI Y， et al. MRI-Based Radiomics Nomogram for Selecting Ovarian Preservation Treatment in Patients With Early-Stage Endometrial Cancer［J］. Front Oncol， 2021， 11： 1-11. doi:10.3389/fonc.2021.730281 doi: 10.3389/fonc.2021.730281
[24]	ZENG Q， LI H， ZHU Y， et al. Development and validation of a predictive model combining clinical， radiomics， and deep transfer learning features for lymph node metastasis in early gastric cancer［J］. Front Med （Lausanne）， 2022， 9： 986437. doi:10.3389/fmed.2022.986437 doi: 10.3389/fmed.2022.986437
[25]	GAO W， WANG W， SONG D， et al. A predictive model integrating deep and radiomics features based on gadobenate dimeglumine-enhanced MRI for postoperative early recurrence of hepatocellular carcinoma［J］. Radiol Med， 2022， 127（3）： 259-271. doi:10.1007/s11547-021-01445-6 doi: 10.1007/s11547-021-01445-6
[26]	GONG J， ZHANG W， HUANG W， et al. CT-based radiomics nomogram may predict local recurrence-free survival in esophageal cancer patients receiving definitive chemoradiation or radiotherapy： A multicenter study［J］. Radiother Oncol， 2022， 174： 8-15. doi:10.1016/j.radonc.2022.06.010 doi: 10.1016/j.radonc.2022.06.010
[27]	AFSHAR P， MOHAMMADI A， PLATANIOTIS K N， et al. From Handcrafted to Deep-Learning-Based Cancer Radiomics： Challenges and opportunities［J］. Ieee Signal Proc Mag， 2019， 36（4）： 132-160. doi:10.1109/msp.2019.2900993 doi: 10.1109/msp.2019.2900993

项目	临床参数标签	影像组学标签	深度迁移学习标签	深度学习影像组学列线图
训练集
AUC	0.850^*	0.878^*	0.920^*	0.958
2.5%CI	0.787	0.824	0.876	0.931
97.5%CI	0.912	0.932	0.963	0.986
准确性	0.808	0.827	0.827	0.904
查准率	0.823	0.849	0.828	0.914
查全率	0.859	0.859	0.891	0.924
F1 Score	0.840	0.854	0.859	0.919
测试集
AUC	0.916	0.709^*	0.842^*	0.951
2.5%CI	0.827	0.539	0.712	0.876
97.5%CI	1.000	0.880	0.972	1.000
准确性	0.872	0.641	0.744	0.923
查准率	0.821	0.680	0.710	0.885
查全率	1.000	0.739	0.957	1.000
F1 Score	0.902	0.708	0.815	0.939

Ultrasound⁃based deep learning radiomics nomogram to differentiate type Ⅰ and type Ⅱ epithelial ovarian cancer

RichHTML

PDF (PC)

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 8

References 27

Related Articles 15

Recommended Articles

Metrics

Comments

[1]	Chengyu QIAN,Linsheng WANG,Xingjiang XIONG,Weidong QIAN. Predictive value of serum LDL⁃C and hs⁃CRP combined with carotid ultrasound parameters in patients with suspected coronary heart disease [J]. The Journal of Practical Medicine, 2025, 41(9): 1401-1406.
[2]	Erfang GUO,Lei FENG,Chaohui SHI,Ning LI,Weiqun LIN,Shuhua. ZHANG. Construction and evaluation of a diagnostic model for female stress urinary incontinence based on the morphology and elasticity of the levator ani muscle by transperineal three⁃dimensional ultrasound combined with shear wave elastography [J]. The Journal of Practical Medicine, 2025, 41(8): 1224-1231.
[3]	Kui DONG,jie WU,Jing YAN,Haitao LIU,Jun WANG,Guan′en. QIAO. Study on risk factors of colorectal adenomatous polyps and construction and validation of prediction model [J]. The Journal of Practical Medicine, 2025, 41(6): 838-845.
[4]	Ying ZOU,Jihua LIU,Jingyi LI,Hai BI,Yan SHI,Xiudi LU,Qibo. ZHANG. Efficacy of transfer learning artificial intelligence model based on ultrasound in evaluating the probability of malignancy of partially cystic thyroid nodule [J]. The Journal of Practical Medicine, 2025, 41(6): 889-895.
[5]	Runfang WANG,Ya DUAN,Liyan DU,Xiaodan LIU,Wenning LIAN,Yan HUO,Dandan. YANG. Epidemiological status and risk factors associated with placental abruption among pregnant women in Hebei Province [J]. The Journal of Practical Medicine, 2025, 41(6): 904-910.
[6]	Dongli LIU,Zilin QUAN,Lingxiu ZHONG,Qiqi CHEN,Wenqiao CAI,Senpei ZHUANG,Ying WEI,Huiyi PAN,Yawen. LIN. Construction and validation of a predictive model for antibiotic-associated diarrhea after surgery in children with congenital heart disease [J]. The Journal of Practical Medicine, 2025, 41(5): 683-690.
[7]	Qiaoying MO,Fangyi ZHU,Cunkui ZHU,Shenglong. MA. Nomogram model of malnutrition risk in patients suffering from chronic heart failure grounded on GNRI score [J]. The Journal of Practical Medicine, 2025, 41(5): 691-698.
[8]	Danhui LAI,Yanhui JIANG,Siting YE,Shulian ZHUANG,Shuang YANG,Wen XUE,Jianxing ZHANG. Analysis of prediction of carotid in-stent restenosis based on ultrasonographic carotid plaque radiomics [J]. The Journal of Practical Medicine, 2025, 41(5): 742-750.
[9]	Changhui WU,Zhiping HUANG,Huiping DAI,Huifang QIU,Chun HE,Fang. TANG. Energy efficiency of contrast⁃enhanced ultrasound combined with TERT promoter mutation to construct a nomogram model for the prediction of concomitant cervical lymph node metastasis in PTMC [J]. The Journal of Practical Medicine, 2025, 41(5): 756-765.
[10]	Wei LIU,Jun HOU,Longquan TANG,Peng ZHOU,Yan ZHONG,Qinyan LUO,Xiaoyu KUANG,Hua LIU,Ziqing XIONG,Wei XIONG,Chenggao WU,Aiping. LE. Analysis of influencing factors of blood transfusion in children with traumatic brain injury and construction of prediction model： A multi⁃center retrospective study [J]. The Journal of Practical Medicine, 2025, 41(4): 553-560.
[11]	Kui XU,Jun. ZHOU. Research progress of ultrasound microbubbles in diagnosis and treatment of thyroid cancer [J]. The Journal of Practical Medicine, 2025, 41(3): 454-458.
[12]	Runhong CUI,Yan CHI,Di ZHANG,Hexiao GAO,Xue. JIANG. Efficacy of ultrasound therapy combined with muscle imbalance adjustment training in patients with adhesive capsulitis oftheshoulder [J]. The Journal of Practical Medicine, 2025, 41(2): 208-214.
[13]	Xinran ZHANG,Yan SHEN,Jiaojiao HU,Qingqing CHEN,Yangjie XIAO,Feng LU,Shasha YUAN,Xiaohong FU. Study on the applied value of combined clinical and ultrasound multiparameter constructed nomogram for predicting HER⁃2⁃positive breast cancer [J]. The Journal of Practical Medicine, 2025, 41(18): 2812-2819.
[14]	Shanshan YAN,Huiyu ZHANG,Jun GU,Kan ZHANG,Yu ZHAO. Value of preoperative ultrasound combined with NLR， Ctn and CEA levels in predicting cervical lymph node metastasis of papillary thyroid carcinoma [J]. The Journal of Practical Medicine, 2025, 41(18): 2859-2864.
[15]	Lingchun CAO,Fanliang MENG,Xiaoan SHENG. Risk factors and model construction of immune therapy⁃related thyroid dysfunction in non⁃small cell lung cancer [J]. The Journal of Practical Medicine, 2025, 41(17): 2705-2714.