Objective To explore the value of constructing a model to predict the number of positive cores in systematic biopsy in prostate cancer (PCa) using a combination of radiomics features based on magnetic resonance imaging and clinical indicators. Methods Retrospectively collected magnetic resonance imaging and clinical data from two medical institutions (Gansu Provincial Hospital from January 2018 to February 2023, Zhangye People's Hospital Affiliated to Hexi College from April 2020 to February 2023). The 155 patients from Gansu Provincial Hospital were randomly divided into a training set (n=109; 80 cases with positive needle count ≥6 and 29 cases with positive needle count <6) and an internal validation set (n=46; 34 cases with positive needle count ≥6 and 12 cases with positive needle count <6) in a 7:3 ratio. The 43 patients from Zhangye People's Hospital Affiliated to Hexi College were used as external validation set. Small field of view high-resolution T2-weighted imaging (sFOV HR-T2WI) and contrast-enhanced delayed-phase images were selected to extract radiomic features from the three-dimensional region of interest of the entire prostate, and radiomics model was constructed and Radscores calculated after dimensionality reduction and feature selection. Univariate and multivariate logistic regressions were used to screen for independent risk factors for positive cores in systematic biopsy. Nomogram was constructed using Radscore and clinical independent risk factors to predict the number of positive cores in systematic biopsy in PCa patients, which was then externally validated. Results Age, alkaline phosphatase (ALP), free prostate specific antigen (FPSA), total prostate specific antigen (TPSA), FPSA/TPSA ratio, and prostate specific antigen density (PSAD) were not statistically significantly different between the training, internal validation, and external validation sets (P>0.05). FPSA, TPSA, FPSA/TPSA ratio, and PSAD were significantly different between the positive cores <6 and positive cores ≥6 groups (P<0.001). Univariate logistic regression analysis showed that FPSA (P<0.001), TPSA (P<0.001), FPSA/TPSA ratio (P=0.001), PSAD (P<0.001), and Radscore (P<0.001) were risk factors for positive cores in systematic biopsy in PCa. Multivariate logistic regression analysis showed that PSAD (OR=0.251, 95%CI 0.063-0.996, P=0.049) and Radscore (OR=1.990, 95%CI 1.409-2.812, P<0.001) were independent risk factors for positive cores in systematic biopsy in PCa. The clinical models achieved AUCs of 0.849(95%CI 0.774-0.924), 0.817(95%CI 0.693-0.941), and 0.631(95%CI 0.439-0.822); the 12 features for radiomics models are derived solely from sFOV HR-T2WI, the radiomics models achieved AUCs of 0.868(95%CI 0.791-0.945), 0.846(95%CI 0.695-0.996), and 0.815(95%CI 0.660-0.970); the nomogram achieved AUCs of 0.921(95%CI 0.869-0.973), 0.868(95%CI 0.743-0.992), and 0.840(95%CI 0.702-0.978) in the training set, internal validation set, and external validation set, respectively. Conclusions The combination of radiomic features extracted from sFOV HR-T2WI and PSAD can preoperatively be used as a noninvasive manner to predict the number of positive cores of the PCa patients. This approach has a certain value in risk stratification of PCa patients and guiding personalized clinical management.
, figureFileSmall=8V4qRMo6ZXMqi6vVfqaLiA==, figureFileBig=SmYhw2WLoRCg4pTjqoncsw==, tableContent=null), ArticleFig(id=1198318901814718758, tenantId=1146029695717560320, journalId=1189873630562394117, articleId=1198196209157505615, language=EN, label=Fig.3, caption=Violin plots showing the differences in Radscore between groups with positive cores ≥6 and positive cores<6 in the training set (A), internal validation set (B), and external validation set (C), figureFileSmall=6GFRh5skWrJqw9/M9s2bwA==, figureFileBig=AfrFUSDvLpkaLtxct57Jqg==, tableContent=null), ArticleFig(id=1198318901902799147, tenantId=1146029695717560320, journalId=1189873630562394117, articleId=1198196209157505615, language=CN, label=图3, caption=影像组学评分在训练集(A)、内部验证集(B)和外部验证集(C)中阳性针数≥6和阳性针数<6组间差异的小提琴图
***P<0.001，****P<0.0001
, figureFileSmall=6GFRh5skWrJqw9/M9s2bwA==, figureFileBig=AfrFUSDvLpkaLtxct57Jqg==, tableContent=null), ArticleFig(id=1198318901990879534, tenantId=1146029695717560320, journalId=1189873630562394117, articleId=1198196209157505615, language=EN, label=Fig.4, caption=Nomogram for predicting positive cores, verification and efficiency evaluation of the nomogram, figureFileSmall=7Xwl0l9drxtzr58gejF2ow==, figureFileBig=1Fmu0ETjLvvfQFWNmLz8mA==, tableContent=null), ArticleFig(id=1198318902057988400, tenantId=1146029695717560320, journalId=1189873630562394117, articleId=1198196209157505615, language=CN, label=图4, caption=预测前列腺癌系统穿刺阳性针数的列线图及其准确性验证和预测效能评价
A. PSAD和Radscore联合构建的预测模型列线图；B、C. 临床模型、影像组学模型和列线图的校准曲线(B)及决策曲线(C)
, figureFileSmall=7Xwl0l9drxtzr58gejF2ow==, figureFileBig=1Fmu0ETjLvvfQFWNmLz8mA==, tableContent=null), ArticleFig(id=1198318902125097268, tenantId=1146029695717560320, journalId=1189873630562394117, articleId=1198196209157505615, language=EN, label=null, caption=ROC curves of the clinical models, radiomics models, and nomogram in the training set (A), internal validation set (B), and external validation set (C), figureFileSmall=0DukKGOyu/eQ0+736doSmg==, figureFileBig=6kWZhHRtGOzZYER2YQedDQ==, tableContent=null), ArticleFig(id=1198318902188011832, tenantId=1146029695717560320, journalId=1189873630562394117, articleId=1198196209157505615, language=CN, label=图5, caption=临床模型、影像组学模型和列线图在训练集(A)、内部验证集(B)和外部验证集(C)中的ROC曲线, figureFileSmall=0DukKGOyu/eQ0+736doSmg==, figureFileBig=6kWZhHRtGOzZYER2YQedDQ==, tableContent=null), ArticleFig(id=1198318902263509309, tenantId=1146029695717560320, journalId=1189873630562394117, articleId=1198196209157505615, language=EN, label=Tab.1, caption=
Objective To explore the value of constructing a model to predict the number of positive cores in systematic biopsy in prostate cancer (PCa) using a combination of radiomics features based on magnetic resonance imaging and clinical indicators. Methods Retrospectively collected magnetic resonance imaging and clinical data from two medical institutions (Gansu Provincial Hospital from January 2018 to February 2023, Zhangye People's Hospital Affiliated to Hexi College from April 2020 to February 2023). The 155 patients from Gansu Provincial Hospital were randomly divided into a training set (n=109; 80 cases with positive needle count ≥6 and 29 cases with positive needle count <6) and an internal validation set (n=46; 34 cases with positive needle count ≥6 and 12 cases with positive needle count <6) in a 7:3 ratio. The 43 patients from Zhangye People's Hospital Affiliated to Hexi College were used as external validation set. Small field of view high-resolution T2-weighted imaging (sFOV HR-T2WI) and contrast-enhanced delayed-phase images were selected to extract radiomic features from the three-dimensional region of interest of the entire prostate, and radiomics model was constructed and Radscores calculated after dimensionality reduction and feature selection. Univariate and multivariate logistic regressions were used to screen for independent risk factors for positive cores in systematic biopsy. Nomogram was constructed using Radscore and clinical independent risk factors to predict the number of positive cores in systematic biopsy in PCa patients, which was then externally validated. Results Age, alkaline phosphatase (ALP), free prostate specific antigen (FPSA), total prostate specific antigen (TPSA), FPSA/TPSA ratio, and prostate specific antigen density (PSAD) were not statistically significantly different between the training, internal validation, and external validation sets (P>0.05). FPSA, TPSA, FPSA/TPSA ratio, and PSAD were significantly different between the positive cores <6 and positive cores ≥6 groups (P<0.001). Univariate logistic regression analysis showed that FPSA (P<0.001), TPSA (P<0.001), FPSA/TPSA ratio (P=0.001), PSAD (P<0.001), and Radscore (P<0.001) were risk factors for positive cores in systematic biopsy in PCa. Multivariate logistic regression analysis showed that PSAD (OR=0.251, 95%CI 0.063-0.996, P=0.049) and Radscore (OR=1.990, 95%CI 1.409-2.812, P<0.001) were independent risk factors for positive cores in systematic biopsy in PCa. The clinical models achieved AUCs of 0.849(95%CI 0.774-0.924), 0.817(95%CI 0.693-0.941), and 0.631(95%CI 0.439-0.822); the 12 features for radiomics models are derived solely from sFOV HR-T2WI, the radiomics models achieved AUCs of 0.868(95%CI 0.791-0.945), 0.846(95%CI 0.695-0.996), and 0.815(95%CI 0.660-0.970); the nomogram achieved AUCs of 0.921(95%CI 0.869-0.973), 0.868(95%CI 0.743-0.992), and 0.840(95%CI 0.702-0.978) in the training set, internal validation set, and external validation set, respectively. Conclusions The combination of radiomic features extracted from sFOV HR-T2WI and PSAD can preoperatively be used as a noninvasive manner to predict the number of positive cores of the PCa patients. This approach has a certain value in risk stratification of PCa patients and guiding personalized clinical management.
Key words
prostate neoplasms
/
radiomics
/
magnetic resonance imaging
/
positive cores
/
prostate specific antigen density
Ni-Ni Pan, Jing Li, Jian-Xin Zhao, Liu-Yan Shi, Lian-Qiu Xiong, Li-Li Ma, Ying-Chao Wang, Lian-Ping Zhao, Gang Huang.
Prediction of the number of positive cores in systematic biopsy of prostate cancer using MRI radiomics combined with clinical indicators[J].
Medical Journal of Chinese People’s Liberation Army,
2024
, 49
(12)
: 1350
-1359
.
DOI: 10.11855/j.issn.0577-7402.1057.2024.0105
正文
收起
前列腺癌(prostatic cancer，PCa)是全球男性第2位常见癌症和第5位癌症死亡原因[1]，其发病率逐年升高[2]。作为一种高度异型性肿瘤[3]，85%的PCa呈多灶性及分散性生长[4]，其精准诊断和评估具有一定的挑战[5]。经直肠超声引导下系统穿刺活检是目前公认的PCa诊断金标准，不仅能够对穿刺组织进行Gleason病理分级，还可通过系统穿刺的阳性针数反映肿瘤的范围和多灶性[6]。中国临床肿瘤学会(Chinese Society of Clinical Oncology，CSCO)PCa诊疗指南2022[7]中指出，≥50%穿刺针数阳性与<50%穿刺针数阳性PCa患者的预后存在明显差异，阳性针数的精准识别有利于患者的个体化医疗。系统穿刺活检具有侵入性，可能引起出血、感染、疼痛、血尿、血精、迷走神经反射、附睾炎等并发症，且存在一定的抽样误差，在一定程度上限制了其临床应用。基于多参数MRI(multi-parameter magnetic resonance，mpMRI)的影像组学模型已被证实能够较好地无创预测PCa的Gleason分级[8]，尤其是基于整个腺体分割图像的影像组学特征训练的模型AUC最高达到了0.87[9]。但目前尚无影像组学模型用于预测PCa系统穿刺阳性针数的研究。为此，本研究探讨基于全前列腺分割的MRI图像中提取的影像组学特征联合临床指标构建模型预测PCa患者系统穿刺阳性针数的价值，以期为PCa患者风险分层及指导临床决策提供帮助。
收集患者的术前临床资料[年龄、血清游离前列腺特异性抗原(free prostate specific antigen，FPSA)、血清总前列腺特异性抗原(total prostate specific antigen，TPSA)、FPSA/TPSA比值、碱性磷酸酶(alkaline phosphatase，ALP)]及病理穿刺结果等。计算前列腺特异性抗原密度(prostate specific antigen density，PSAD)：PSAD[ng/(ml.cm3)]=TPSA(ng/ml)/前列腺体积(cm3)，其中前列腺体积(cm3)=最大左右径(cm)×最大前后径(cm)×最大上下径(cm)×0.52。
两家医院MRI扫描均采用3.0T MRI(德国西门子公司)。所有患者常规MRI扫描序列包括T1加权成像(T1-weighted imaging，T1WI)、T2加权成像(T2-weighted imaging，T2WI)、弥散加权成像(diffusion-weighted imaging，DWI)、小视野高分辨率T2加权成像(small field of view high-resolution T2-weighted imaging，sFOV HR-T2WI)和动态对比增强(dynamic contrast enhancement，DCE)序列。本研究所采用序列的详细参数如表1所示。
1.5 图像处理与分割
将sFOV HR-T2WI和增强扫描延迟期图像以DICOM格式导入ITK-SNAP软件，勾画感兴趣区(region of interest，ROI)，并将原始图像和相对应ROI以NiFti格式保存。由1名工作经验为5年的放射科医师沿前列腺轮廓逐层勾画边界，避开患者尿道、射精管、精阜及精囊根部，勾画完成后软件自动生成全前列腺三维感兴趣容积(volum of interest，VOI)。随机抽取30例患者，由同一名医师在3个月后再次进行ROI勾画，用于评估影像组学特征的再现性和稳定性。
1.6 特征提取与选择
对分割后的图像进行N4偏置场校正处理，消除磁场不均匀对信号强度的影响。将全前列腺VOI导入FeAture Explorer(v0.5.5；https://github.com/salan668/FAE)软件，对图像进行等体素重采样，体素大小为1 mm×1 mm×1 mm，图像体素强度值通过固定箱宽度(binwidth)=25进行离散化处理，体素强度范围归一化至[0，1]之间，以减少不同机型及扫描参数对影像组学特征的影响。提取一阶、形态、灰度共生矩阵(gray level co-occurrence matrix，GLCM)、灰度游程矩阵(gray level run length matrix，GLRLM)、灰度大小区域矩阵(gray level size zone matrix，GLSZM)、灰度相关矩阵(gray level dependence matrix，GLDM)和邻域灰度差矩阵(neighbouring gray tone difference matrix，NGTDM)参数影像组学特征。所有影像组学特征均遵循成像生物标志物标准化倡议(Imaging Biomarker Standardization Initiative，IBSI)指南[10]。采用合成少数类过采样技术(synthetic minority oversampling technique，SMOTE)平衡样本数据。对所有特征进行Z-score标准化处理，用皮尔森相关系数(Pearson correlation coefficients，PCC)对特征进行降维处理，系数阈值设定为0.99。最后，使用方差分析(analysis of variance，ANOVA)、Relief算法、递归式特征消除(recursive feature elimination，RFE)及Kruskal-Wallis(K-W测试)进行特征筛选。选择的分类器有支持向量机(support vector machine，SVM)、线性判别分析(linear discriminant analisis，LDA)、逻辑回归(logistic regression，LR)和LASSO逻辑回归(logistic regression via LASSO，LR-LASSO)。训练集模型验证采用内部5-折交叉验证设置参数。
1.7 模型构建与效能评估
通过单因素及多因素logistic回归分析筛选系统穿刺阳性针数的独立危险因素；将筛选的临床独立危险因素及组学特征采用logistic回归分别建立临床模型、影像组学模型及列线图。通过受试者工作特征(receiver operating characteristic，ROC)曲线下面积(area under the curve，AUC)对模型的性能进行评估。根据最大约登指数界值计算准确度(accuracy)、敏感度(sensitivity)、特异度(specificity)、阳性预测值(positive predictive value，PPV)和阴性预测值(negative predictive value，NPV)。用Bootstrape重采样法估计1000个样本的95%置信区间(95%CI)。使用Hosmer-Lemeshow检验和校准曲线评估列线图的拟合程度，采用ROC和决策曲线分析(decision curve analysis，DCA)衡量模型的鉴别效能和临床应用价值。以上所有过程使用Feature Explorer Pro(FAE，v0.5.5)、R(v4.3.0)、GraphPad Prism(v9.0.0)及IBM SPSS Statistics 21.0软件实现。
Global Cancer Statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries
SungH, FerlayJ, SiegelRL, et al. Global Cancer Statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J]. CA Cancer J Clin, 2021, 71(3): 209-249.
[2]
Recent global patterns in prostate cancer incidence and mortality rates
CulpMB, SoerjomataramI, EfstathiouJA, et al. Recent global patterns in prostate cancer incidence and mortality rates[J]. Eur Urol, 2020, 77(1): 38-52.
[3]
Inter- and intra-tumor heterogeneity of metastatic prostate cancer determined by digital spatial gene expression profiling
BradyL, KrinerM, ColemanI, et al. Inter- and intra-tumor heterogeneity of metastatic prostate cancer determined by digital spatial gene expression profiling[J]. Nat Commun, 2021, 12(1): 1426.
Comments on national guidelines for diagnosis and treatment of prostate cancer 2022 in China (English version)
YangY. Comments on national guidelines for diagnosis and treatment of prostate cancer 2022 in China (English version)[J]. Chin J Cancer Res, 2022, 34(5): 456-457.
[8]
Radiomics features measured with multiparametric magnetic resonance imaging predict prostate cancer aggressiveness
HectorsSJ, ChernyM, YadavKK, et al. Radiomics features measured with multiparametric magnetic resonance imaging predict prostate cancer aggressiveness[J]. J Urol, 2019, 202(3): 498-505.
[9]
Prediction of prostate cancer disease aggressiveness using bi-parametric MRI radiomics
RodriguesA, SantinhaJ, GalvãoB, et al. Prediction of prostate cancer disease aggressiveness using bi-parametric MRI radiomics[J]. Cancers (Basel), 2021, 13(23): 6065.
[10]
The image biomarker standardization initiative: standardized quantitative radiomics for high-throughput image-based phenotyping
ZwanenburgA, VallièresM, AbdalahMA, et al. The image biomarker standardization initiative: standardized quantitative radiomics for high-throughput image-based phenotyping[J]. Radiology, 2020, 295(2): 328-338.
Magnetic resonance imaging-defined prostate-specific antigen density significantly improves the risk prediction for clinically significant prostate cancer on biopsy
BhatNR, VetterJM, AndrioleGL, et al. Magnetic resonance imaging-defined prostate-specific antigen density significantly improves the risk prediction for clinically significant prostate cancer on biopsy[J]. Urology, 2019, 126: 152-157.
[14]
PSA density in the diagnosis of prostate cancer in the Chinese population: results from the Chinese Prostate Cancer Consortium
SongZJ, QianJK, YangY, et al. PSA density in the diagnosis of prostate cancer in the Chinese population: results from the Chinese Prostate Cancer Consortium[J]. Asian J Androl, 2021, 23(3): 300-305.
Utility of T2-weighted MRI texture analysis in assessment of peripheral zone prostate cancer aggressiveness: a single-arm, multicenter study
NketiahGA, ElschotM, ScheenenTW, et al. Utility of T2-weighted MRI texture analysis in assessment of peripheral zone prostate cancer aggressiveness: a single-arm, multicenter study[J]. Sci Rep, 2021, 11(1): 2085.
[17]
Prediction of clinically significant cancer using radiomics features of pre-biopsy of multiparametric MRI in men suspected of prostate cancer
OgbonnayaCN, ZhangX, AlsaediBSO, et al. Prediction of clinically significant cancer using radiomics features of pre-biopsy of multiparametric MRI in men suspected of prostate cancer[J]. Cancers (Basel), 2021, 13(24): 6199.
[18]
Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach
AertsHJWL, VelazquezER, LeijenaarRTH, et al. Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach[J]. Nat Commun, 2014, 5: 4006.
[19]
MRI-based radiomics signature: a potential biomarker for identifying glypican 3-positive hepatocellular carcinoma
GuD, XieY, WeiJ, et al. MRI-based radiomics signature: a potential biomarker for identifying glypican 3-positive hepatocellular carcinoma[J]. J Magn Reson Imaging, 2020, 52(6): 1679-1687.
[20]
A wavelet features derived radiomics nomogram for prediction of malignant and benign early-stage lung nodules
JingR, WangJ, LiJ, et al. A wavelet features derived radiomics nomogram for prediction of malignant and benign early-stage lung nodules[J]. Sci Rep, 2021, 11(1): 22330.
[21]
CT-based radiomics analysis of different machine learning models for differentiating benign and malignant parotid tumors
ZhengY, ZhouD, LiuH, et al. CT-based radiomics analysis of different machine learning models for differentiating benign and malignant parotid tumors[J]. Eur Radiol, 2022, 32(10): 6953-6964.
[22]
Multiparametric-MRI-based radiomics model for differentiating primary central nervous system lymphoma from glioblastoma: development and cross-vendor validation
XiaW, HuB, LiH, et al. Multiparametric-MRI-based radiomics model for differentiating primary central nervous system lymphoma from glioblastoma: development and cross-vendor validation[J]. J Magn Reson Imaging, 2021, 53(1): 242-250.
[23]
Multiparametric magnetic resonance imaging-based peritumoral radiomics for preoperative prediction of the presence of extracapsular extension with prostate cancer
BaiH, XiaW, JiX, et al. Multiparametric magnetic resonance imaging-based peritumoral radiomics for preoperative prediction of the presence of extracapsular extension with prostate cancer[J]. J Magn Reson Imaging, 2021, 54(4): 1222-1230.
[24]
Multiparametric MRI for prostate cancer characterization: combined use of radiomics model with PI-RADS and clinical parameters
WoźnickiP, WesthoffN, HuberT, et al. Multiparametric MRI for prostate cancer characterization: combined use of radiomics model with PI-RADS and clinical parameters[J]. Cancers (Basel), 2020, 12(7): 1767.
[25]
Machine learning classifiers can predict Gleason pattern 4 prostate cancer with greater accuracy than experienced radiologists
AntonelliM, JohnstonEW, DikaiosN, et al. Machine learning classifiers can predict Gleason pattern 4 prostate cancer with greater accuracy than experienced radiologists[J]. Eur Radiol, 2019, 29(9): 4754-4764.
[26]
MRI-based radiomics nomogram for predicting prostate cancer with gray-zone prostate-specific antigen levels to reduce unnecessary biopsies
ZhangL, ZhangJ, TangM, et al. MRI-based radiomics nomogram for predicting prostate cancer with gray-zone prostate-specific antigen levels to reduce unnecessary biopsies[J]. Diagnostics (Basel), 2022, 12(12): 3005.
[27]
MRI based radiomics compared with the PI-RADS V2.1 in the prediction of clinically significant prostate cancer: biparametric vs multiparametric MRI
ChenT, ZhangZ, TanS, et al. MRI based radiomics compared with the PI-RADS V2.1 in the prediction of clinically significant prostate cancer: biparametric vs multiparametric MRI[J]. Front Oncol, 2021, 11: 792456.
[28]
Comparison of biparametric and multiparametric MRI in the diagnosis of prostate cancer
XuL, ZhangG, ShiB, et al. Comparison of biparametric and multiparametric MRI in the diagnosis of prostate cancer[J]. Cancer Imaging, 2019, 19(1): 90.
[29]
Comparison of biparametric and multiparametric MRI for clinically significant prostate cancer detection with PI-RADS version 2.1
TamadaT, KidoA, YamamotoA, et al. Comparison of biparametric and multiparametric MRI for clinically significant prostate cancer detection with PI-RADS version 2.1[J]. J Magn Reson Imaging, 2021, 53(1): 283-291.
Multiparametric MRI-based radiomics model to predict pelvic lymph node invasion for patients with prostate cancer
ZhengH, MiaoQ, LiuY, et al. Multiparametric MRI-based radiomics model to predict pelvic lymph node invasion for patients with prostate cancer[J]. Eur Radiol, 2022, 32(8): 5688-5699.
[32]
Predicting tumor perineural invasion status in high-grade prostate cancer based on a clinical-radiomics model incorporating T2-weighted and diffusion-weighted magnetic resonance images
ZhangW, ZhangW, LiX, et al. Predicting tumor perineural invasion status in high-grade prostate cancer based on a clinical-radiomics model incorporating T2-weighted and diffusion-weighted magnetic resonance images[J]. Cancers (Basel), 2022, 15(1): 86.
[33]
Magnetic resonance imaging radiomics-based machine learning prediction of clinically significant prostate cancer in equivocal PI-RADS 3 lesions
HectorsSJ, ChenC, ChenJ, et al. Magnetic resonance imaging radiomics-based machine learning prediction of clinically significant prostate cancer in equivocal PI-RADS 3 lesions[J]. J Magn Reson Imaging, 2021, 54(5): 1466-1473.
[34]
Cellular landscape of tumour microenvironment in prostate cancer
FengD, XiongQ, WeiQ, et al. Cellular landscape of tumour microenvironment in prostate cancer[J]. Immunology, 2023, 168(2): 199-202.
[35]
CT radiomics, radiologists, and clinical information in predicting outcome of patients with COVID-19 pneumonia
HomayouniehF, EbrahimianS, BabaeiR, et al. CT radiomics, radiologists, and clinical information in predicting outcome of patients with COVID-19 pneumonia[J]. Radiol Cardiothorac Imaging, 2020, 2(4): e200322.
[36]
The potential of prostate gland radiomic features in identifying the Gleason score
GongL, XuM, FangM, et al. The potential of prostate gland radiomic features in identifying the Gleason score[J]. Comput Biol Med, 2022, 144: 105318.
[37]
Exploring the predictive value of additional peritumoral regions based on deep learning and radiomics: a multicenter study
WuX, DongD, ZhangL, et al. Exploring the predictive value of additional peritumoral regions based on deep learning and radiomics: a multicenter study[J]. Med Phys, 2021, 48(5): 2374-2385.
第 1 页 / 共 1 页
2024年第49卷第12期
PDF下载
183
78
引用本文
BibTeX
文章信息
doi: 10.11855/j.issn.0577-7402.1057.2024.0105
接收时间：2023-08-09
首发时间：2025-11-20
出版时间：2024-12-28
补充材料
相关文章
文章信息
作者
出版历史
收稿日期：2023-08-09
录用日期：2023-09-06
基金
Beijing Medical Award Foundation(YXJL-2022-0665-0197)
北京医学奖励基金会项目(YXJL-2022-0665-0197)
Grant from the Gansu Provincial Hospital(22GSSYD-33)
甘肃省人民医院院内科研基金项目(22GSSYD-33)
University Innovation Fund Project of Education Department of Gansu Province(2021B-258)
Fig.3 Violin plots showing the differences in Radscore between groups with positive cores ≥6 and positive cores<6 in the training set (A), internal validation set (B), and external validation set (C)
***P<0.001，****P<0.0001
图4 预测前列腺癌系统穿刺阳性针数的列线图及其准确性验证和预测效能评价
Fig.4 Nomogram for predicting positive cores, verification and efficiency evaluation of the nomogram
A. PSAD和Radscore联合构建的预测模型列线图；B、C. 临床模型、影像组学模型和列线图的校准曲线(B)及决策曲线(C)
ROC curves of the clinical models, radiomics models, and nomogram in the training set (A), internal validation set (B), and external validation set (C)
PDF下载
全屏
183
78
关闭全屏
北京市海淀区学院南路86号
010-62199257
qkjq@cast.org.cn
