Sarah Englander

A Combined Architectural and Kinetic Interpretation Model for Breast MR Images

RATIONALE AND OBJECTIVES The purpose of this study was to integrate contrast material kinetic and architectural data from magnetic resonance (MR) images and to assess the improvement in diagnostic accuracy. MATERIALS AND METHODS MR imaging data from a diagnostic cohort of 100 patients (50 malignant and 50 benign cases) were analyzed. RESULTS Qualitative classification of the enhancement curve was the most predictive kinetic feature. Receiver operating characteristic (ROC) curves were calculated for the architectural model alone and for the architectural model combined with the qualitative kinetic classification. The results demonstrated a statistically significant increase in ROC area (P = .03) of the combined model compared with that of the architectural model alone. CONCLUSION The addition of qualitative classification of the time-signal intensity curve to an architectural interpretation model results in significant improvement in model performance as measured by the area under the ROC curve.

Textural Kinetics: A Novel Dynamic Contrast-Enhanced (DCE)-MRI Feature for Breast Lesion Classification

Dynamic contrast-enhanced (DCE)-magnetic resonance imaging (MRI) of the breast has emerged as an adjunct imaging tool to conventional X-ray mammography due to its high detection sensitivity. Despite the increasing use of breast DCE-MRI, specificity in distinguishing malignant from benign breast lesions is low, and interobserver variability in lesion classification is high. The novel contribution of this paper is in the definition of a new DCE-MRI descriptor that we call textural kinetics, which attempts to capture spatiotemporal changes in breast lesion texture in order to distinguish malignant from benign lesions. We qualitatively and quantitatively demonstrated on 41 breast DCE-MRI studies that textural kinetic features outperform signal intensity kinetics and lesion morphology features in distinguishing benign from malignant lesions. A probabilistic boosting tree (PBT) classifier in conjunction with textural kinetic descriptors yielded an accuracy of 90%, sensitivity of 95%, specificity of 82%, and an area under the curve (AUC) of 0.92. Graph embedding, used for qualitative visualization of a low-dimensional representation of the data, showed the best separation between benign and malignant lesions when using textural kinetic features. The PBT classifier results and trends were also corroborated via a support vector machine classifier which showed that textural kinetic features outperformed the morphological, static texture, and signal intensity kinetics descriptors. When textural kinetic attributes were combined with morphologic descriptors, the resulting PBT classifier yielded 89% accuracy, 99% sensitivity, 76% specificity, and an AUC of 0.91.

STEP: Spatiotemporal enhancement pattern for MR-based breast tumor diagnosis

The authors propose a spatiotemporal enhancement pattern (STEP) for comprehensive characterization of breast tumors in contrast-enhanced MR images. By viewing serial contrast-enhanced MR images as a single spatiotemporal image, they formulate the STEP as a combination of (1) dynamic enhancement and architectural features of a tumor, and (2) the spatial variations of pixelwise temporal enhancements. Although the latter has been widely used by radiologists for diagnostic purposes, it has rarely been employed for computer-aided diagnosis. This article presents two major contributions. First, the STEP features are introduced to capture temporal enhancement and its spatial variations. This is essentially carried out through the Fourier transformation and pharmacokinetic modeling of various temporal enhancement features, followed by the calculation of moment invariants and Gabor texture features. Second, for effectively extracting the STEP features from tumors, we develop a graph-cut based segmentation algorithm that aims at refining coarse manual segmentations of tumors. The STEP features are assessed through their diagnostic performance for differentiating between benign and malignant tumors using a linear classifier (along with a simple ranking-based feature selection) in a leave-one-out cross-validation setting. The experimental results for the proposed features exhibit superior performance, when compared to the existing approaches, with the area under the ROC curve approaching 0.97.

Computerized Image Analysis for Identifying Triple-Negative Breast Cancers and Differentiating Them from Other Molecular Subtypes of Breast Cancer on Dynamic Contrast-enhanced MR Images: A Feasibility Study

PURPOSE To determine the feasibility of using a computer-aided diagnosis (CAD) system to differentiate among triple-negative breast cancer, estrogen receptor (ER)-positive cancer, human epidermal growth factor receptor type 2 (HER2)-positive cancer, and benign fibroadenoma lesions on dynamic contrast material-enhanced (DCE) magnetic resonance (MR) images. MATERIALS AND METHODS This is a retrospective study of prospectively acquired breast MR imaging data collected from an institutional review board-approved, HIPAA-compliant study between 2002 and 2007. Written informed consent was obtained from all patients. The authors collected DCE MR images from 65 women with 76 breast lesions who had been recruited into a larger study of breast MR imaging. The women had triple-negative (n = 21), ER-positive (n = 25), HER2-positive (n = 18), or fibroadenoma (n = 12) lesions. All lesions were classified as Breast Imaging Reporting and Data System category 4 or higher on the basis of previous imaging. Images were subject to quantitative feature extraction, feed-forward feature selection by means of linear discriminant analysis, and lesion classification by using a support vector machine classifier. The area under the receiver operating characteristic curve (Az) was calculated for each of five lesion classification tasks involving triple-negative breast cancers. RESULTS For each pair-wise lesion type comparison, linear discriminant analysis helped identify the most discriminatory features, which in conjunction with a support vector machine classifier yielded an Az of 0.73 (95% confidence interval [CI]: 0.59, 0.87) for triple-negative cancer versus all non-triple-negative lesions, 0.74 (95% CI: 0.60, 0.88) for triple-negative cancer versus ER- and HER2-positive cancer, 0.77 (95% CI: 0.63, 0.91) for triple-negative versus ER-positive cancer, 0.74 (95% CI: 0.58, 0.89) for triple-negative versus HER2-positive cancer, and 0.97 (95% CI: 0.91, 1.00) for triple-negative cancer versus fibroadenoma. CONCLUSION Triple-negative cancers possess certain characteristic features on DCE MR images that can be captured and quantified with CAD, enabling good discrimination of triple-negative cancers from non-triple-negative cancers, as well as between triple-negative cancers and benign fibroadenomas. Such CAD algorithms may provide added diagnostic benefit in identifying the highly aggressive triple-negative cancer phenotype with DCE MR imaging in high-risk women.

Segmentation and classification of breast tumor using dynamic contrast-enhanced MR images

Accuracy of automatic cancer diagnosis is largely determined by two factors, namely, the precision of tumor segmentation, and the suitability of extracted features for discrimination between malignancy and benignancy. In this paper, we propose a new framework for accurate characterization of tumors in contrast enhanced MR images. First, a new graph cut based segmentation algorithm is developed for refining coarse manual segmentation, which allows precise identification of tumor regions. Second, by considering serial contrast-enhanced images as a single spatio-temporal image, a spatio-temporal model of segmented tumor is constructed to extract Spatio-Temporal Enhancement Patterns (STEPs). STEPs are designed to capture not only dynamic enhancement and architectural features, but also spatial variations of pixel-wise temporal enhancement of the tumor. While temporal enhancement features are extracted through Fourier transform, the resulting STEP framework captures spatial patterns of temporal enhancement features via moment invariants and rotation invariant Gabor textures. High accuracy of the proposed framework is a direct consequence of this two pronged approach, which is validated through experiments yielding, for instance, an area of 0.97 under the ROC curve.

De-enhancing the dynamic contrast-enhanced breast MRI for robust registration

Dynamic enhancement causes serious problems for registration of contrast enhanced breast MRI, due to variable uptakes of agent on different tissues or even same tissues in the breast. We present an iterative optimization algorithm to de-enhance the dynamic contrast-enhanced breast MRI and then register them for avoiding the effects of enhancement on image registration. In particular, the spatially varying enhancements are modeled by a Markov Random Field, and estimated by a locally smooth function with boundaries using a graph cut algorithm. The de-enhanced images are then registered by conventional B-spline based registration algorithm. These two steps benefit from each other and are repeated until the results converge. Experimental results show that our two-step registration algorithm performs much better than conventional mutual information based registration algorithm. Also, the effects of tumor shrinking in the conventional registration algorithms can be effectively avoided by our registration algorithm.

Reproducibility of visual activation in functional magnetic resonance imaging at very high field strength (4 Tesla).

PURPOSE The reproducibility of functional magnetic resonance imaging (fMRI) has been studied on 1.5 Tesla (T) (high field strength) scanners. We report the reproducibility of visual activation in fMRI at 4 T (very high field strength). METHODS Five healthy subjects were scanned twice in the same session with a 4 T scanner during binocular flashing visual stimulation. The activated areas during the first and second acquisition were compared. RESULTS Activation of the visual cortex was observed in all subjects and activation of lateral geniculate nucleus was also detected in four subjects. The ratio of overlapping activated voxels in the first and second acquisition was 0.81 +/- 0.05. CONCLUSIONS Reproducibility of visual activation using fMRI at 4 T was found to be acceptable, and the results from 4T scanners show a reliability similar to those at 1.5 T.

Phase 1 and pharmacodynamic trial of everolimus in combination with cetuximab in patients with advanced cancer

Preclinical and clinical studies suggest mTOR (mammalian target of rapamycin) inhibitors may have metabolic and antiangiogenic effects, and synergize with epidermal growth factor pathway inhibitors. Therefore, a phase 1/pharmacodynamic trial of everolimus with cetuximab was performed.

Automatic Coil Selection for Streak Artifact Reduction in Radial MRI

In radial MR imaging, streaking artifacts contaminating the entire field of view can arise from regions at the outer edges of the prescribed field of view. This can occur even when the Nyquist criterion is satisfied within the desired field of view. These artifacts become exacerbated when parts of the object lie in the superior/inferior regions of the scanner where the gradient strengths become weakened. When multiple coil arrays are used for signal reception, coils at the outer edges can be disabled before data acquisition to reduce the artifact levels. However, as the weakened gradient strengths near the edges often distort the object, causing the signal to become highly concentrated into a small region, the streaks are often not completely removed. Data from certain coils can also be excluded during reconstruction by visually inspecting the individual coil images, but this is impractical for routine use. In this work, a postprocessing method is proposed to automatically identify those coils whose images contain high levels of streaking for subsequent exclusion during reconstruction. The proposed method was demonstrated in vivo dynamic contrast enhanced MRI datasets acquired using a three‐dimensional hybrid radial sequence. The results demonstrate that the proposed strategy substantially improves the image quality and show excellent agreement with images reconstructed with manually determined coil selection. Magn Reson Med, 2012.

Functional Magnetic Resonance Imaging of Eye Dominance at 4 Tesla

We studied eye dominance in visual cortex and lateral geniculate nucleus (LGN) using functional magnetic resonance imaging (fMRI) at a very high magnetic field (4 tesla). Eight normal volunteers were studied with fMRI at 4 tesla during alternating monocular visual stimulation. The acquisition was repeated twice in 4 subjects to confirm reproducibility. In addition, magnetic resonance signal intensities during three conditions (right eye stimulation, left eye stimulation, and control condition) were compared to determine whether the observed area was truly or relatively monocular in 2 subjects. In both the individual and group analyses, the anterior striate cortex was consistently activated by the contralateral eye more than the ipsilateral eye. Additionally, we found evidence that there were areas in the bilateral LGN which were more active during the stimulation of the contralateral eye than during the stimulation of the ipsilateral eye. The activated areas were reproducible, and the mean ratio of the overlapping area was 0.71 for the repeated scans. The additional experiment revealed that the area in the anterior visual cortex could be divided into two parts, one truly monocular and the other relatively monocular. Our finding confirmed previous fMRI results at 1.5 tesla showing that eye dominance was observed in the contralateral anterior visual cortex. However, the eye dominance in the visual cortex was found not only in the most anterior area corresponding to the monocular temporal crescent but also in the more posterior area, presumably showing the greater sensitivity of the temporal visual field (nasal retina) as compared with the nasal visual field (temporal retina) in the peripheral visual field (peripheral retina). In addition, it is suggested that the nasotemporal asymmetry of the retina and the visual fields is represented in the LGN as well as in the visual cortex.