Dataset Information

Quantitative analysis of vascular properties derived from ultrafast DCE-MRI to discriminate malignant and benign breast tumors.

ABSTRACT: PURPOSE:We propose a novel methodology to integrate morphological and functional information of tumor-associated vessels to assist in the diagnosis of suspicious breast lesions. THEORY AND METHODS:Ultrafast, fast, and high spatial resolution DCE-MRI data were acquired on 15 patients with suspicious breast lesions. Segmentation of the vasculature from the surrounding tissue was performed by applying a Hessian filter to the enhanced image to generate a map of the probability for each voxel to belong to a vessel. Summary measures were generated for vascular morphology, as well as the inputs and outputs of vessels physically connected to the tumor. The ultrafast DCE-MRI data was analyzed by a modified Tofts model to estimate the bolus arrival time, Ktrans (volume transfer coefficient), and vp (plasma volume fraction). The measures were compared between malignant and benign lesions via the Wilcoxon test, and then incorporated into a logistic ridge regression model to assess their combined diagnostic ability. RESULTS:A total of 24 lesions were included in the study (13 malignant and 11 benign). The vessel count, Ktrans , and vp showed significant difference between malignant and benign lesions (P = 0.009, 0.034, and 0.010, area under curve [AUC] = 0.76, 0.63, and 0.70, respectively). The best multivariate logistic regression model for differentiation included the vessel count and bolus arrival time (AUC = 0.91). CONCLUSION:This study provides preliminary evidence that combining quantitative characterization of morphological and functional features of breast vasculature may provide an accurate means to diagnose breast cancer.

SUBMITTER: Wu C

PROVIDER: S-EPMC6347496 | biostudies-literature | 2019 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Quantitative analysis of vascular properties derived from ultrafast DCE-MRI to discriminate malignant and benign breast tumors.

Wu Chengyue C Pineda Federico F Hormuth David A DA Karczmar Gregory S GS Yankeelov Thomas E TE

Magnetic resonance in medicine 20181028 3

<h4>Purpose</h4>We propose a novel methodology to integrate morphological and functional information of tumor-associated vessels to assist in the diagnosis of suspicious breast lesions.<h4>Theory and methods</h4>Ultrafast, fast, and high spatial resolution DCE-MRI data were acquired on 15 patients with suspicious breast lesions. Segmentation of the vasculature from the surrounding tissue was performed by applying a Hessian filter to the enhanced image to generate a map of the probability for eac ...[more]

PMID: 30368906

Similar Datasets

Project description:BackgroundComputer-aided methods have been widely applied to diagnose lesions detected on breast MRI, but fully-automatic diagnosis using deep learning is rarely reported.PurposeTo evaluate the diagnostic accuracy of mass lesions using region of interest (ROI)-based, radiomics and deep-learning methods, by taking peritumor tissues into consideration.Study typeRetrospective.PopulationIn all, 133 patients with histologically confirmed 91 malignant and 62 benign mass lesions for training (74 patients with 48 malignant and 26 benign lesions for testing).Field strength/sequence3T, using the volume imaging for breast assessment (VIBRANT) dynamic contrast-enhanced (DCE) sequence.Assessment3D tumor segmentation was done automatically by using fuzzy-C-means algorithm with connected-component labeling. A total of 99 texture and histogram parameters were calculated for each case, and 15 were selected using random forest to build a radiomics model. Deep learning was implemented using ResNet50, evaluated with 10-fold crossvalidation. The tumor alone, smallest bounding box, and 1.2, 1.5, 2.0 times enlarged boxes were used as inputs.Statistical testsThe malignancy probability was calculated using each model, and the threshold of 0.5 was used to make a diagnosis.ResultsIn the training dataset, the diagnostic accuracy was 76% using three ROI-based parameters, 84% using the radiomics model, and 86% using ROI + radiomics model. In deep learning using the per-slice basis, the area under the receiver operating characteristic (ROC) was comparable for tumor alone, smallest and 1.2 times box (AUC = 0.97-0.99), which were significantly higher than 1.5 and 2.0 times box (AUC = 0.86 and 0.71, respectively). For per-lesion diagnosis, the highest accuracy of 91% was achieved when using the smallest bounding box, and that decreased to 84% for tumor alone and 1.2 times box, and further to 73% for 1.5 times box and 69% for 2.0 times box. In the independent testing dataset, the per-lesion diagnostic accuracy was also the highest when using the smallest bounding box, 89%.Data conclusionDeep learning using ResNet50 achieved a high diagnostic accuracy. Using the smallest bounding box containing proximal peritumor tissue as input had higher accuracy compared to using tumor alone or larger boxes.Level of evidence3 Technical Efficacy: Stage 2.

Project description:The purpose was to analyze the value of quantitative parameters of DCE-MRI in evaluating micro-infiltration of malignant bone tumors.MethodsThirty-nine New Zealand white rabbits were used to establish malignant bone tumor models by implanting VX2 tumor fragments into the right tibiae. After three weeks, models were examined by conventional MRI and DCE-MRI; then the right tibiae were cut into sagittal sections and partitioned into histology slices for comparison with microscopic findings. Micro-infiltration groups were selected and the range of infiltration was determined under the microscope, and corresponding DCE images analyzed to obtain the quantitative parameters include Ktrans, Kep, ve and vp in parenchyma areas, micro-infiltration areas and simple edema areas. One-way ANOVA was used to compare the differences of the parameters between the three areas. Receiver operating characteristic curves (ROCs) were plotted to determine the accuracy of different parameters by area under curves (AUCs).Results22 cases (22/39, 56.4%) were included in the micro-infiltration group and the infiltration depth ranged from 1.3 mm to 4.6 mm, with an average depth of 3.2 mm ± 0.8 mm. The statistical results of quantitative parameters in the three areas were as follows: Ktrans values were (0.494 ± 0.052), (0.403 ± 0.049), (0.173 ± 0.047) min-1 (p = =0.000), Kep values were (1.959 ± 0.65), (1.528 ± 0.372), (1.174 ± 0.486) min-1 (p = =0.000), ve values were (0.247 ± 0.068), (0.283 ± 0.057), (0.168 ± 0.062) min-1 (p = =0.000), vp values were (0.125 ± 0.036), (0.108 ± 0.033), (0.098 ± 0.025) min-1 (p = =0.022), respectively. Ktrans and Kep values had significant difference in the three areas after comparing between-groups, respectively. However, there were no significant difference in vp values between parenchyma and micro-infiltration areas (p = =0.078), micro-infiltration and simple edema areas (p = =0.315), and ve values between parenchyma and micro-infiltration areas (p = =0.056). The ve values were higher in parenchyma and micro-infiltration areas then simple edema areas. Ktrans had highest accuracy in differentiating different areas (AUC > 0.9), respectively.ConclusionQuantitative parameters Ktrans, Kep and ve can assess the extent of intramedullary invasion of malignant bone tumors. Ktrans have highest accuracy in differentiating different regions.

Project description:Purpose:To find potentially diagnostic texture analysis (TA) features and to evaluate the diagnostic accuracy of two-dimensional (2D) magnetic resonance (MR) TA for differentiation between hepatocellular carcinoma (HCC) and benign hepatocellular tumors in the non-cirrhotic liver in an exploratory MR-study. Materials and methods:108 non-cirrhotic patients (62 female; 41.5 ± 18.3 years) undergoing preoperative contrast-enhanced MRI were retrospectively included in this multi-center-study. TA including gray-level histogram, co-occurrence and run-length matrix features (total 19 features) was performed by two independent readers. Native fat-saturated-T1w and T2w as well as arterial and portal-venous post contrast-enhanced 2D-image-slices were assessed. Conventional reading was performed by two separate independent readers. Differences in TA features between HCC and benign lesions were investigated using independent sample t-tests. Logistic regression analysis was performed to obtain the optimal number/combination of TA-features and diagnostic accuracy of TA analysis. Sensitivity and specificity of the better performing radiologist were compared to TA analysis. Results:The highest number of significantly differing TA-features (n = 5) was found using the arterial-phase images including one gray-level histogram (skewness, p = 0.018) and four run-length matrix features (all, p < 0.02). The optimal binary logistic regression model for TA-features of the arterial-phase images contained 13 parameters with an accuracy of 84.5% (sensitivity 84.1%, specificity 84.9%) and area-under-the-curve of 0.92 (95%-confidence-interval 0.85-0.98) for diagnosis of HCC. Conventional reading yielded a significantly lower sensitivity (63.6%, p = 0.027) and no significant difference in specificity (94.6%, p = 0.289) at best. Conclusion:2D-TA of MR images is a feasible objective method that may help to distinguish HCC from benign hepatocellular tumors in the non-cirrhotic liver. Most promising results were found in TA features in the arterial phase images.

Project description:PURPOSE:The aim of this study was to develop a portable perfusion phantom and validate its utility in quantitative dynamic contrast-enhanced magnetic resonance imaging of the abdomen. METHODS:A portable perfusion phantom yielding a reproducible contrast enhancement curve (CEC) was developed. A phantom package including perfusion and static phantoms were imaged simultaneously with each of three healthy human volunteers in two different 3T MR scanners. Look-up tables correlating reference (known) contrast concentrations with measured ones were created using either the static or perfusion phantom. Contrast maps of image slices showing four organs (liver, spleen, pancreas, and paravertebral muscle) were generated before and after data correction using the look-up tables. The contrast concentrations at 4.5 min after dosing in each of the four organs were averaged for each volunteer. The mean contrast concentrations (4 organs × 3 volunteers = 12) were compared for the two scanners, and the intra-class correlation coefficient (ICC) was calculated. Also, the ICC of the mean Ktrans values between the two scanners was calculated before and after data correction. RESULTS:The repeatability coefficient of CECs of perfusion phantom was higher than 0.997 in all measurements. The ICC of the tissue contrast concentrations between the two scanners was 0.693 before correction, but increased to 0.974 after correction using the look-up tables (LUTs) of perfusion phantom. However, the ICC was not increased after correction using static phantom (ICC: 0.617). Similarly, the ICC of the Ktrans values was 0.899 before correction, but increased to 0.996 after correction using perfusion phantom LUTs. The ICC of the Ktrans values, however, was not increased when static phantom LUTs were used (ICC: 0.866). CONCLUSIONS:The perfusion phantom reduced variability in quantitating contrast concentration and Ktrans values of human abdominal tissues across different MR units, but static phantom did not. The perfusion phantom has the potential to facilitate multi-institutional clinical trials employing quantitative DCE-MRI to evaluate various abdominal malignancies.

Project description:The categorization of intraductal proliferative lesions of the breast based on routine light microscopic examination of histopathologic sections is in many cases challenging, even for experienced pathologists. The development of computational tools to aid pathologists in the characterization of these lesions would have great diagnostic and clinical value. As a first step to address this issue, we evaluated the ability of computational image analysis to accurately classify DCIS and UDH and to stratify nuclear grade within DCIS. Using 116 breast biopsies diagnosed as DCIS or UDH from the Massachusetts General Hospital (MGH), we developed a computational method to extract 392 features corresponding to the mean and standard deviation in nuclear size and shape, intensity, and texture across 8 color channels. We used L1-regularized logistic regression to build classification models to discriminate DCIS from UDH. The top-performing model contained 22 active features and achieved an AUC of 0.95 in cross-validation on the MGH data-set. We applied this model to an external validation set of 51 breast biopsies diagnosed as DCIS or UDH from the Beth Israel Deaconess Medical Center, and the model achieved an AUC of 0.86. The top-performing model contained active features from all color-spaces and from the three classes of features (morphology, intensity, and texture), suggesting the value of each for prediction. We built models to stratify grade within DCIS and obtained strong performance for stratifying low nuclear grade vs. high nuclear grade DCIS (AUC = 0.98 in cross-validation) with only moderate performance for discriminating low nuclear grade vs. intermediate nuclear grade and intermediate nuclear grade vs. high nuclear grade DCIS (AUC = 0.83 and 0.69, respectively). These data show that computational pathology models can robustly discriminate benign from malignant intraductal proliferative lesions of the breast and may aid pathologists in the diagnosis and classification of these lesions.

Project description:PurposeTo develop a disposable point-of-care portable perfusion phantom (DP4) and validate its clinical utility in a multi-institutional setting for quantitative dynamic contrast-enhanced magnetic resonance imaging (qDCE-MRI).MethodsThe DP4 phantom was designed for single-use and imaged concurrently with a human subject so that the phantom data can be utilized as the reference to detect errors in qDCE-MRI measurement of human tissues. The change of contrast-agent concentration in the phantom was measured using liquid chromatography-mass spectrometry. The repeatability of the contrast enhancement curve (CEC) was assessed with five phantoms in a single MRI scanner. Five healthy human subjects were recruited to evaluate the reproducibility of qDCE-MRI measurements. Each subject was imaged concurrently with the DP4 phantom at two institutes using three 3T MRI scanners from three different vendors. Pharmacokinetic (PK) parameters in the regions of liver, spleen, pancreas, and paravertebral muscle were calculated based on the Tofts model (TM), extended Tofts model (ETM), and shutter speed model (SSM). The reproducibility of each PK parameter over three measurements was evaluated with the intraclass correlation coefficient (ICC) and compared before and after DP4-based error correction.ResultsThe contrast-agent concentration in the DP4 phantom was linearly increased over 10 min (0.17 mM/min, measurement accuracy: 96%) after injecting gadoteridol (100 mM) at a constant rate (0.24 ml/s, 4 ml). The repeatability of the CEC within the phantom was 0.997 when assessed by the ICC. The reproducibility of the volume transfer constant, Ktrans , was the highest of the PK parameters regardless of the PK models. The ICCs of Ktrans in the TM, ETM, and SSM before DP4-based error correction were 0.34, 0.39, and 0.72, respectively, while those increased to 0.93, 0.98, and 0.86, respectively, after correction.ConclusionsThe DP4 phantom is reliable, portable, and capable of significantly improving the reproducibility of qDCE-MRI measurements.

Dataset Information

Quantitative analysis of vascular properties derived from ultrafast DCE-MRI to discriminate malignant and benign breast tumors.

Publications

Quantitative analysis of vascular properties derived from ultrafast DCE-MRI to discriminate malignant and benign breast tumors.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets