ABSTRACT
Objective@#To investigate the prognostic value of radiomics analysis in predicting axillary lymph nodes (ALN) metastasis of breast cancer based on dynamic contrast-enhanced MR imaging (DCE-MRI).@*Methods@#One hundred and ninety-six patients with suspected breast cancer were prospectively collected for dynamic breast DCE-MRI. Enhanced MR imaging data of 72 axillary lymph nodes were evaluated separately by a chief radiologist and a resident, and the consistency analysis was performed. Lymph nodes were dichotomized according to the pathology results derived from operation or biopsy under real-time virtual sonography based on MRI data. Clinical and imaging data were also divided into corresponding groups. (Imaging) Data from both groups were respectively classified as training set and testing set by stratified sampling in proportion with 3∶1. AK software was applied to extract 6 major categories of 385 features (including histogram, morphology, texture parameters, gray level co-occurrence matrix, run-length matrix and grey level zone size matrix from imaging), and a set of statistically significant features were subsequently obtained by dimension reduction. The prediction model was established through binary classification logistic regression and employed to externally test the validation set by the method of confusion matrix. Meanwhile, ROC analysis was applied to assess the diagnostic performance of the model.@*Results@#Of the 72 axillary lymph nodes, 35 were metastatic negative and 37 were positive. The consistency of enhanced MRI radiomics features was good, between 0.841 and 0.980. Uniformity, ClusterProminence_AllDirection_offset1_SD, Correlation_AllDirection_offset1, LongRunEmphasis_angle90_offset7 and SurfaceVolumeRatio were statistically significant differences (P<0.01), the area under the ROC between 0.747 and 0.931. In the training and testing group, the areas under the ROC, sensitivity, specificity and accuracy of the model were 0.953, 0.893, 0.926, 92.6% (50/54) and 0.944, 0.900, 1.000, 88.9% (16/18) respectively.@*Conclusion@#The prediction model based on radiomic features may provide a non-invasive and effective approach to the assessment of the risk of ALN metastasis of breast cancer.
ABSTRACT
Objective To investigate the prognostic value of radiomics analysis in predicting axillary lymph nodes (ALN) metastasis of breast cancer based on dynamic contrast-enhanced MR imaging (DCE-MRI). Methods One hundred and ninety-six patients with suspected breast cancer were prospectively collected for dynamic breast DCE-MRI. Enhanced MR imaging data of 72 axillary lymph nodes were evaluated separately by a chief radiologist and a resident, and the consistency analysis was performed. Lymph nodes were dichotomized according to the pathology results derived from operation or biopsy under real-time virtual sonography based on MRI data. Clinical and imaging data were also divided into corresponding groups. (Imaging) Data from both groups were respectively classified as training set and testing set by stratified sampling in proportion with 3∶1. AK software was applied to extract 6 major categories of 385 features (including histogram, morphology, texture parameters, gray level co-occurrence matrix, run-length matrix and grey level zone size matrix from imaging), and a set of statistically significant features were subsequently obtained by dimension reduction. The prediction model was established through binary classification logistic regression and employed to externally test the validation set by the method of confusion matrix. Meanwhile, ROC analysis was applied to assess the diagnostic performance of the model. Results Of the 72 axillary lymph nodes, 35 were metastatic negative and 37 were positive. The consistency of enhanced MRI radiomics features was good, between 0.841 and 0.980. Uniformity, ClusterProminence_AllDirection_offset1_SD, Correlation_AllDirection_offset1, LongRunEmphasis_angle90_offset7 and SurfaceVolumeRatio were statistically significant differences (P<0.01), the area under the ROC between 0.747 and 0.931. In the training and testing group, the areas under the ROC, sensitivity, specificity and accuracy of the model were 0.953, 0.893, 0.926, 92.6% (50/54) and 0.944, 0.900, 1.000, 88.9% (16/18) respectively. Conclusion The prediction model based on radiomic features may provide a non-invasive and effective approach to the assessment of the risk of ALN metastasis of breast cancer.
ABSTRACT
Objective To provide objective parameters for differentiating pancreatic cystic tumors via using computed tomography texture analysis (CTTA) to quantify the special imaging features of pancreatic cystadenomas.Methods Enhanced CT images of pancreas from patients who were admitted in Department of Radiology in First Hospital affiliated with Zhejiang University and First People's Hospital of Hangzhou City and pathologically diagnosed as pancreatic serous cystadenomas (n =48) and mucinous cystadenomas (n =34) from January 2009 to December 2016 were retrospectively analyzed.Regions of interest were drawn on the parenchymal phase CT images in 5 slices according to the border of the tumors.Mean grey level intensity (M),variance (V),entropy (E),skewness (Ske) and kurtosis (Kur) were obtained from fine texture (σ =1.0) to coarse texture (σ =2.5).Receiver operating characteristic (ROC) curve for texture parameters with statistically difference was drawn,and the area under curve (AUC),diagnostic sensitivity and specificity were calculated.The diagnostic accuracy of senior and junior doctors was compared with the traditional CT analysis method.Results Reliability coefficient of the two radiologists was 0.809 ～ 0.997 with high consistency.Compared with mucinous cystic tumors,serous cystadenomas had a significantly different V (5.93 ± 9.02 vs 1.29±0.62),E (2.39±0.61 vs2.02±0.39) and Kur(30.18 ±42.55 vs 8.80-±4.34) in Ske0 of 2.5 (P ＜0.05),and there were no statistically significant differences on other parameters.The AUC of differential parameters for diagnosing two kinds of cystic tumors ranged from 0.56 to 0.84.The diagnostic accuracy of the traditional CT analysis method by junior doctor and senior radiologist was 60％ and 71％,respectively.Conclusions CTTA can not only effectively quantify the heterogeneity of pancreatic cystadenomas,but also is effective in the differentiation.
ABSTRACT
Objective Implement random sample from a simulation population,to evaluate the The impact of samplesize and sample-process on several usual importance evaluate methods,observe the stability of those methods.Methods This study introduced existed importance methods,using PROC SURVEYSELECT procedure to sample a fixed population for 1000 times,generating 1000 same size sample,to evaluate the stability of relative importance methods.We sampled the population to generate datasets with different sample size to observe impact of sample-size on those methods.Results The sum of squared correlation coefficients' estimator is bigger than model R-square,squared standardized regression coefficients' sum is smaller.In contrary,sum of the Product Measure,Relative Weight and Dominance Analysis are extremely close to model R-square.When the sample size small than 1000,the estimator have obviously variation,but the variation decreased when the sample size rise up.Conclusion The dominance analysis has best stability,also has the best match of model R2 in those methods.