Development and validation of peritumoral vascular and intratumoral radiomics to predict pathologic complete responses to neoadjuvant chemotherapy in patients with triple-negative breast cancer.

BACKGROUND: To develop and validate a peritumoral vascular and intratumoral radiomics model to improve pretreatment predictions for pathologic complete responses (pCRs) to neoadjuvant chemoradiotherapy (NAC) in patients with triple-negative breast cancer (TNBC). METHODS: A total of 282 TNBC patients (93 in the primary cohort, 113 in the validation cohort, and 76 in The Cancer Imaging Archive [TCIA] cohort) were retrospectively included. The peritumoral vasculature on the maximum intensity projection (MIP) from pretreatment DCE-MRI was segmented by a Hessian matrix-based filter and then edited by a radiologist. Radiomics features were extracted from the tumor and peritumoral vasculature of the MIP images. The LASSO method was used for feature selection, and the k-nearest neighbor (k-NN) classifier was trained and validated to build a predictive model. The diagnostic performance was assessed using the ROC analysis. RESULTS: One hundred of the 282 patient (35.5%) with TNBC achieved pCRs after NAC. In predicting pCRs, the combined peritumoral vascular and intratumoral model (fusion model) yields a maximum AUC of 0.82 (95% confidence interval [CI]: 0.75, 0.88) in the primary cohort, a maximum AUC of 0.67 (95% CI: 0.57, 0.76) in the internal validation cohort, and a maximum AUC of 0.65 (95% CI: 0.52, 0.78) in TCIA cohort. The fusion model showed improved performance over the intratumoral model and the peritumoral vascular model, but not significantly (p > 0.05). CONCLUSION: This study suggested that combined peritumoral vascular and intratumoral radiomics model could provide a non-invasive tool to enable prediction of pCR in TNBC patients treated with NAC.

Quantitative analysis from ultrafast dynamic contrast-enhanced breast MRI using population-based versus individual arterial input functions, and comparison with semi-quantitative analysis.

PURPOSE: To evaluate the value of inline quantitative analysis of ultrafast dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) using a population-based arterial input function (P-AIF) compared with offline quantitative analysis with an individual AIF (I-AIF) and semi-quantitative analysis for diagnosing breast cancer. METHODS: This prospective study included 99 consecutive patients with 109 lesions (85 malignant and 24 benign). Model-based parameters (Ktrans, kep, and ve) and model-free parameters (washin and washout) were derived from CAIPIRINHA-Dixon-TWIST-VIBE (CDTV) DCE-MRI. Univariate analysis and multivariate logistic regression analysis with forward stepwise covariate selection were performed to identify significant variables. The AUC and F1 score were assessed for semi-quantitative and two quantitative analyses. RESULTS: kep from inline quantitative analysis with P-AIF for diagnosing breast cancer provided an AUC similar to kep from offline quantitative analysis with I-AIF (0.782 vs 0.779, p = 0.954), higher compared to washin from semi-quantitative analysis (0.782 vs 0.630, p = 0.034). Furthermore, the inline quantitative analysis with P-AIF achieved the larger F1 score (0.920) compared with offline quantitative analysis with I-AIF (0.780) and semi-quantitative analysis (0.480). There were no statistically significant differences for kep values between the two quantitative analysis schemes (p = 0.944). CONCLUSION: The inline quantitative analysis with P-AIF from CDTV in characterizing breast lesions could offer similar diagnostic accuracy to offline quantitative analysis with I-AIF, and higher diagnostic accuracy to semi-quantitative analysis.

Model-Free and Model-based Parameters Derived From CAIPIRINHA-Dixon-TWIST-VIBE DCE-MRI: Associations With Prognostic Factors and Molecular Subtypes of Invasive Ductal Breast Cancer.

BACKGROUND: CAIPIRINHA-Dixon-TWIST-VIBE (CDTV) dynamic contrast-enhanced MRI (DCE-MRI) can be used to characterize breast cancer. However, the influence of the clinicopathologic factors and molecular subtypes of invasive breast carcinoma (IDC) on the model-free and model-based parameters has not been investigated. PURPOSE: To compare model-free and model-based parameters of CDTV DCE-MRI with both clinicopathologic factors and molecular subtypes of IDC. STUDY TYPE: Prospective. POPULATION: A total of 152 patients (mean age, 52 years) with IDC including 42 luminal A, 64 luminal B, 22 human epidermal growth factor receptor-2 (HER2) positive, and 24 triple-negative subtypes. FIELD STRENGTH/SEQUENCE: A 3 T; turbo-FLASH, Dixon VIBE, and CDTV. ASSESSMENT: Model-free parameters (initial enhancement rate [IER] and maximum slope [MS]) were estimated from the time-intensity curve. The mean, minimum, maximum, and range between the minimum and maximum values of inline model-based parameters (Ktrans , kep , and ve ) were measured to assess intratumoral heterogeneity of IDC lesions. STATISTICAL TESTS: Student's t tests, Mann-Whitney U tests, Kruskal-Wallis tests, post hoc Steel-Dwass tests, and receiver operating characteristic (ROC) curves. P < 0.05 was considered significant. RESULTS: No significant differences in IER and MS values were seen among the clinicopathologic factors and molecular subtypes (Bonferroni-corrected P = 0.011-0.862, P = 0.145-0.601, respectively). The minimum kep values in HER2-positive IDC were significantly lower than those in HER2-negative IDC. The mean and range kep values were independent predictors for distinguishing the high (grade 3) and low (grade 1 or 2) nuclear grade groups according to multivariable analyses. The post hoc test showed that the kep minimum and kep range values were significantly different between luminal A and HER2-positive tumor subtypes, yielding an area-under-the-curve of 0.820. DATA CONCLUSION: Compared with the model-free parameters, inline kep related model-based parameters on CDTV DCE-MRI can be applied as a feasible tool to differentiate luminal A from HER2-positive breast cancers. EVIDENCE LEVEL: 2. TECHNICAL EFFICACY: Stage 2.

Improved value of whole-lesion histogram analysis on DCE parametric maps for diagnosing small breast cancer (≤ 1 cm).

OBJECTIVES: To determine if whole-lesion histogram analysis on dynamic contrast-enhanced (DCE) parametric maps help to improve the diagnostic accuracy of small suspicious breast lesions (≤ 1 cm). METHODS: This retrospective study included 99 female patients with 114 lesions (40 malignant and 74 benign lesions) suspicious on magnetic resonance imaging (MRI).Two radiologists reviewed all lesions and descripted the morphologic and kinetic characteristics according to BI-RADS by consensus. Whole lesions were segmented on DCE parametric maps (washin and washout), and quantitative histogram features were extracted. Univariate analysis and multivariate logistic regression analysis with forward stepwise covariate selection were performed to identify significant variables. Diagnostic performance was assessed and compared with that of qualitative BI-RADS assessment and quantitative histogram analysis by ROC analysis. RESULTS: For malignancy defined as a washout or plateau pattern, the qualitative kinetic pattern showed a significant difference between the two groups (p = 0.023), yielding an AUC of 0.603 (95% confidence interval [CI]: 0.507, 0.694). The mean and median of washout were independent quantitative predictors of malignancy (p = 0.002, 0.010), achieving an AUC of 0.796 (95% CI: 0. 709, 0.865). The AUC of the quantitative model was better than that of the qualitative model (p < 0.001). CONCLUSIONS: Compared with the qualitative BI-RADS assessment, quantitative whole-lesion histogram analysis on DCE parametric maps was better to discriminate between small benign and malignant breast lesions (≤ 1 cm) initially defined as suspicious on DCE-MRI. KEY POINTS: • For malignancy defined as a washout or plateau, the kinetic pattern may provide information to diagnose small breast cancer. • The mean and median of washout map were significantly lower for small malignant breast lesions than for benign lesions. • Quantitative histogram analysis on MRI parametric maps improves diagnostic accuracy for small breast cancer, which may obviate unnecessary biopsy.

Differentiation between idiopathic granulomatous mastitis and invasive breast carcinoma, both presenting with non-mass enhancement without rim-enhanced masses: The value of whole-lesion histogram and texture analysis using apparent diffusion coefficient.

PURPOSE: The aim of this study was to investigate whether whole-lesion histogram and texture analysis using apparent diffusion coefficient can discriminate between idiopathic granulomatous mastitis (IGM) and invasive breast carcinoma (IBC), both of which appeared as non-mass enhancement lesions without rim-enhanced masses. METHOD: This retrospective study included 58 pathology-proven female patients at two independent study sites (27 IGM patients and 31 IBC patients). Diffusion-weighted imaging (3b values, 50, 400 or 500, and 800 s/mm2) was performed using 1.5 T or 3 T MR scanners from the same vendor. Whole-lesions were segmented and 11 features were extracted. Univariate analysis and multivariate logistic regression analysis were performed to identify significant variables for differentiating IGM from IBC. Receiver operating characteristic curve was assessed. The interobserver reliability between two observers for the histogram and texture measurement was also reported. RESULTS: The 5th percentile, difference entropy and entropy of apparent diffusion coefficient showed significant differences between the two groups. An area under the curve of 0.778 (95 % CI: 0.648, 0.908), accuracy of 79.3 %, and sensitivity of 87.1 % was achieved using these three significant features. No significant feature was found with the multivariate analysis. For the interobserver reliability, all apparent diffusion coefficient parameters except skewness and kurtosis indicated good or excellent agreement, while these two features showed moderate agreement. CONCLUSIONS: Whole-lesion histogram and texture analysis using apparent diffusion coefficient provide a non-invasive analytical approach to the differentiation between IGM and IBC, both presenting with non-mass enhancement without rim-enhanced masses.

Machine Learning-Based Analysis of MR Multiparametric Radiomics for the Subtype Classification of Breast Cancer.

Objective: To investigate whether machine learning analysis of multiparametric MR radiomics can help classify immunohistochemical (IHC) subtypes of breast cancer. Study design: One hundred and thirty-four consecutive patients with pathologically-proven invasive ductal carcinoma were retrospectively analyzed. A total of 2,498 features were extracted from the DCE and DWI images, together with the new calculated images, including DCE images changing over six time points (DCEsequential) and DWI images changing over three b-values (DWIsequential). We proposed a novel two-stage feature selection method combining traditional statistics and machine learning-based methods. The accuracies of the 4-IHC classification and triple negative (TN) vs. non-TN cancers were assessed. Results: For the 4-IHC classification task, the best accuracy of 72.4% was achieved based on linear discriminant analysis (LDA) or subspace discrimination of assembled learning in conjunction with 20 selected features, and only small dependent emphasis of Kendall-tau-b for sequential features, based on the DWIsequential with the LDA model, yielding an accuracy of 53.7%. The linear support vector machine (SVM) and medium k-nearest neighbor using eight features yielded the highest accuracy of 91.0% for comparing TN to non-TN cancers, and the maximum variance for DWIsequential alone, together with a linear SVM model, achieved an accuracy of 83.6%. Conclusions: Whole-tumor radiomics on MR multiparametric images, DCE images changing over time points, and DWI images changing over different b-values provide a non-invasive analytical approach for breast cancer subtype classification and TN cancer identification.

Differentiation of triple-negative breast cancer from other subtypes through whole-tumor histogram analysis on multiparametric MR imaging.

PURPOSE: To identify triple-negative (TN) breast cancer imaging biomarkers in comparison to other molecular subtypes using multiparametric MR imaging maps and whole-tumor histogram analysis. MATERIALS AND METHODS: This retrospective study included 134 patients with invasive ductal carcinoma. Whole-tumor histogram-based texture features were extracted from a quantitative ADC map and DCE semi-quantitative maps (washin and washout). Univariate analysis using the Student's t test or Mann-Whitney U test was performed to identify significant variables for differentiating TN cancer from other subtypes. The ROC curves were generated based on the significant variables identified from the univariate analysis. The AUC, sensitivity, and specificity for subtype differentiation were reported. RESULTS: The significant parameters on the univariate analysis achieved an AUC of 0.710 (95% confidence interval [CI] 0.562, 0.858) with a sensitivity of 63.6% and a specificity of 73.1% at the best cutoff point for differentiating TN cancers from Luminal A cancers. An AUC of 0.763 (95% CI 0.608, 0.917) with a sensitivity of 86.4% and a specificity of 72.2% was achieved for differentiating TN cancers from human epidermal growth factor receptor 2 (HER2) positive cancers. Also, an AUC of 0.683 (95% CI 0.556, 0.809) with a sensitivity of 54.5% and a specificity of 83.9% was achieved for differentiating TN cancers from non-TN cancers. There was no significant feature on the univariate analysis for TN cancers versus Luminal B cancers. CONCLUSIONS: Whole-tumor histogram-based imaging features derived from ADC, along with washin and washout maps, provide a non-invasive analytical approach for discriminating TN cancers from other subtypes. KEY POINTS: • Whole-tumor histogram-based features on MR multiparametric maps can help to assess biological characterization of breast cancer. • Histogram-based texture analysis may predict the molecular subtypes of breast cancer. • Combined DWI and DCE evaluation helps to identify triple-negative breast cancer.

Quantitative analysis of neovascular permeability in glioma by dynamic contrast-enhanced MR imaging.

This study was designed to quantitatively analyse neovascular permeability in glioma by dynamic contrast-enhanced MRI (DCE-MRI). Forty-four patients with glioma were included in this study. The highest value of volume transfer constant (K(trans)) and volume of extravascular extracellular space per unit volume of tissue (V(e)) were obtained and the differences in K(trans) and V(e) between low-grade glioma (LGG) and high-grade glioma (HGG) were analyzed. The correlations between K(trans), V(e) and glioma grade were performed. Receiver operating characteristic (ROC) curve analyses were conducted. The values of K(trans) and V(e) of LGG were significantly lower than those of HGG. The correlation analysis demonstrated statistically significant relationships between K(trans) and glioma grade, between V(e) and glioma grade, and between K(trans) and V(e). The ROC curve analyses of K(trans) (0.035/min) and V(e) (0.130) for differentiating LGG from HGG were statistically significant. Thus, DCE-MRI can be used to estimate neovascular permeability and for pre-operative grading of glioma.

[Medical image retrieval by high level semantic features and low level content features of image].

Content-based image retrieval aims at searching the similar images using low level features,and medical image retrieval needs it for the retrieval of similar images. Medical images contain not only a lot of content data, but also a lot of semantic information. This paper presents an approach by combining digital imaging and communications in medicine (DICOM) features and low level features to perform retrieval on medical image databases. At the first step, the semantic information is extracted from DICOM header for the pre-filtering of the images, and then dual-tree complex wavelet transfrom(DT-CWT) features of pre-filtered images and example images are extracted to retrieve similar images. Experimental results show that by combining the high level semantics (DICOM features) and low level content features (texture) the retrieval time is reduced and the performance of medical image retrieval is increased.