Deepa Sheth

Bronchial Artery Embolization

Hemoptysis represents a significant clinical entity with high morbidity and potential mortality. Most hemorrhages from a bronchial source arise in the setting of chronic inflammatory diseases. Medical management (in terms of resuscitation and bronchoscopic interventions) and surgery have severe limitations in these patient populations. Embolization procedures represent the first-line treatment for hemoptysis arising from a bronchial arterial source. This article discusses anatomical and technical considerations, as well as outcomes and complications, in the setting of bronchial arterial embolization in the treatment of hemoptysis.

Value of breast MRI for patients with a biopsy showing atypical ductal hyperplasia (ADH)

To evaluate the diagnostic value of dynamic contrast‐enhanced (DCE) magnetic resonance imaging (MRI) for patients with atypical ductal hyperplasia (ADH) in predicting malignant upgrade.

Fast bilateral breast coverage with high spectral and spatial resolution (HiSS) MRI at 3T

To develop and assess a full‐coverage, sensitivity encoding (SENSE)‐accelerated breast high spatial and spectral resolution (HiSS) magnetic resonance imaging (MRI) within clinically reasonable times as a potential nonenhanced MRI protocol for breast density measurement or breast cancer screening.

Transfer Learning From Convolutional Neural Networks for Computer-Aided Diagnosis: A Comparison of Digital Breast Tomosynthesis and Full-Field Digital Mammography

RATIONALE AND OBJECTIVES With the growing adoption of digital breast tomosynthesis (DBT) in breast cancer screening, we compare the performance of deep learning computer-aided diagnosis on DBT images to that of conventional full-field digital mammography (FFDM). MATERIALS AND METHODS In this study, we retrospectively collected FFDM and DBT images of 78 biopsy-proven lesions from 76 patients. A region of interest was selected for each lesion on FFDM, synthesized 2D, and DBT key slice images. Features were extracted from each lesion using a pretrained convolutional neural network (CNN) and served as input to a support vector machine classifier trained in the task of predicting likelihood of malignancy. RESULTS From receiver operating characteristic (ROC) analysis of all 78 lesions, the synthesized 2D image performed best in both the cradiocaudal view (area under the ROC curve [AUC] = 0.81, SE = 0.05) and mediolateral oblique view (AUC = 0.88, SE = 0.04) in the task of lesion characterization. When cradiocaudal and mediolateral oblique data of each lesion were merged through soft voting, DBT key slice image performed best (AUC = 0.89, SE = 0.04). When only masses and architectural distortions (ARDs) were considered, DBT performed significantly better than FFDM (p = 0.024). CONCLUSION DBT performed significantly better than FFDM in the merged view classification of mass and ARD lesions. The increased performance suggests that the information extracted by the CNN from DBT images may be more relevant to lesion malignancy status than the information extracted from FFDM images. Therefore, this study provides supporting evidence for the efficacy of computer-aided diagnosis on DBT in the evaluation of mass and ARD lesions.

Abbreviated MRI and Accelerated MRI for Screening and Diagnosis of Breast Cancer

Abstract Although published studies have revealed that magnetic resonance imaging (MRI) is by far the most effective imaging modality for cancer detection, it is currently considered cost-ineffective for screening women at an intermediate risk for breast cancer. The concept of an “abbreviated MRI” protocol has recently emerged as a possible solution for reducing the cost of MRI. The abbreviated MRI is a shortened version of the standard MRI, consisting of a single early phase dynamic contrast enhanced (DCE) series. Several clinical studies have shown that this MRI protocol would not affect sensitivity or specificity for breast MRI screening purposes. In clinical practice, morphologic evaluation and kinetic assessment are 2 major components of the interpretation process. However, kinetic assessment cannot be performed with the abbreviated protocol, because multiple sets of post-contrast images are necessary for the generation of kinetic curves. “Accelerated MRI” is a collective term for imaging techniques that acquire DCE-MR images in a very short time. Published studies suggest that the kinetic assessment during the very early post-contrast phase obtained with the accelerated MRI techniques is comparable to that with the standard MRI techniques. Applying accelerated MR techniques could potentially enhance the abbreviated MRI protocol in terms of diagnostic potential, while maintaining the shorter study time. Thus, the abbreviated MRI protocol associated with accelerated MRI techniques may provide value for screening and for diagnostic purposes.

Transfer learning with convolutional neural networks for lesion classification on clinical breast tomosynthesis.

With growing adoption of digital breast tomosynthesis (DBT) in breast cancer screening protocols, it is important to compare the performance of computer-aided diagnosis (CAD) in the diagnosis of breast lesions on DBT images compared to conventional full-field digital mammography (FFDM). In this study, we retrospectively collected FFDM and DBT images of 78 lesions from 76 patients, each containing lesions that were biopsy-proven as either malignant or benign. A square region of interest (ROI) was placed to fully cover the lesion on each FFDM, DBT synthesized 2D images, and DBT key slice images in the cranial-caudal (CC) and mediolateral-oblique (MLO) views. Features were extracted on each ROI using a pre-trained convolutional neural network (CNN). These features were then input to a support vector machine (SVM) classifier, and area under the ROC curve (AUC) was used as the figure of merit. We found that in both the CC view and MLO view, the synthesized 2D image performed best (AUC = 0.814, AUC = 0.881 respectively) in the task of lesion characterization. Small database size was a key limitation in this study, and could lead to overfitting in the application of the SVM classifier. In future work, we plan to expand this dataset and to explore more robust deep learning methodology such as fine-tuning.

Diagnostic value of electric properties tomography (EPT) for differentiating benign from malignant breast lesions: comparison with standard dynamic contrast-enhanced MRI

ObjectivesTo evaluate the diagnostic utility of electric properties tomography (EPT) in differentiating benign from malignant breast lesions in comparison with dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI).MethodsIn this institutional review board-approved retrospective study, 116 consecutive patients with 141 breast lesions (50 benign and 91 malignant) underwent 3-T MRI, including 3D turbo-spin echo (TSE) sequence and standard DCE-MRI scans between January 2014 and January 2017. The lesions were segmented semi-automatically using subtraction DCE-MR images, and they were registered to the phase images from 3D TSE. The mean conductivity of the lesion was obtained from phase-based reconstruction of lesions. From the DCE-MRI, initial enhancement rate (IER) and signal enhancement ratio (SER) were calculated from signal intensity (SI) as follows: IER = (SIearly - SIpre)/SIpre, SER = (SIearly - SIpre)/(SIdelayed - SIpre). The parameters from EPT and the DCE-MRI were compared between benign and malignant lesions.ResultsThere was significant difference in mean conductivity (0.14 ± 1.77 vs 1.14 ± 1.36 S/m, p < 0.0001) and SER (0.77 ± 0.28 vs 1.04 ± 0.25, p < 0.0001) between benign and malignant lesions, but not in IER (p = 0.06). Receiver operating curve (ROC) analysis revealed that the area under the curve (AUC) of the mean conductivity and SER was 0.71 and 0.80, respectively, without significant difference (p = 0.15).ConclusionsThe mean conductivity of EPT was significantly different between benign and malignant breast lesions as well as kinetic parameter or SER from DCE-MRI.Key Points• The conductivity of malignant lesions was higher than that of benign lesions.• EPT helps differentiatie benign from malignant lesions.• Diagnostic ability of EPT was not significantly different from that of DCE-MRI.

Lymph node wire localization post-chemotherapy: Towards improving the false negative sentinel lymph node biopsy rate in breast cancer patients

PURPOSE To evaluate whether the disease status of the pre-neoadjuvant chemotherapy (NAC) core biopsied lymph node (preNACBxLN) in patients with node positive breast cancer corresponds to nodal status of all surgically retrieved lymph nodes (LNs) post-NAC and whether wire localization of this LN is feasible. MATERIALS AND METHODS HIPPA compliant IRB approved retrospective study including breast cancer patients (a.) with preNACBxLN confirmed metastases, (b.) who received NAC, and (c.) underwent wire localization of the preNACBxLN. Electronic medical records were reviewed. Fishers exact test was used to compare differences in residual disease post-NAC among breast cancer subtypes. RESULTS 28 women with node positive breast cancer underwent ultrasound guided wire localization of the preNACBxLN, without complication. There was no evidence of residual nodal disease for 16 patients, with mean 4.4 (median 4) LNs resected. 12 patients had residual nodal metastases, with mean 9.2 (median 7) LNs resected and mean 2.3 (median 2) LNs with tumor involvement. 11 patients had metastases detected within the localized LN. One patient had micrometastasis in a sentinel LN, despite no residual disease in the preNACBxLN. Patients with luminal A/B breast cancer more often had residual nodal metastases (86%) at pathology, as compared to patients with HER2+ (20%) and Triple Negative breast cancer (50%), though not quite achieving statistical significance (p=0.055). CONCLUSION Ultrasound guided wire localization of the preNACBxLN is feasible and may improve detection of residual tumor in patients post-NAC.

Deep learning in computer-aided diagnosis incorporating mammographic characteristics of both tumor and parenchyma stroma.

We investigated the additive role of breast parenchyma stroma in the computer-aided diagnosis (CADx) of tumors on full-field digital mammograms (FFDM) by combining images of the tumor and contralateral normal parenchyma information via deep learning. The study included 182 breast lesions in which 106 were malignant and 76 were benign. All FFDM images were acquired using a GE 2000D Senographe system and retrospectively collected under an Institution Review Board (IRB) approved, Health Insurance Portability and Accountability Act (HIPAA) compliant protocol. Convolutional neutral networks (CNNs) with transfer learning were used to extract image-based characteristics of lesions and of parenchymal patterns (on the contralateral breast) directly from the FFDM images. Classification performance was evaluated and compared between analysis of only tumors and that of combined tumor and parenchymal patterns in the task of distinguishing between malignant and benign cases with the area under the Receiver Operating Characteristic (ROC) curve (AUC) used as the figure of merit. Using only lesion image data, the transfer learning method yielded an AUC value of 0.871 (SE=0.025) and using combined information from both lesion and parenchyma analyses, an AUC value of 0.911 (SE=0.021) was observed. This improvement was statistically significant (p-value=0.0362). Thus, we conclude that using CNNs with transfer learning to combine extracted image information of both tumor and parenchyma may improve breast cancer diagnosis.