Benign versus malignant solid breast masses: US differentiation

Differential diagnosis of breast mucinous carcinoma with an oval shape from fibroadenoma based on ultrasonographic features

BACKGROUND

Approximately 50% of breast mucinous carcinomas (MCs) are oval and have the possibility of being misdiagnosed as fibroadenomas (FAs). We aimed to identify the key features that can help differentiate breast MC with an oval shape from FA on ultrasonography (US).

METHODS

Seventy-six MCs from 71 consecutive patients and 50 FAs with an oval shape from 50 consecutive patients were included in our study. All lesions pathologically diagnosed. According to the Breast Imaging Reporting and Data System (BI-RADS), first, the ultrasonographic features of the MCs and FAs were recorded and a final category was assessed. Then, the differences in ultrasonographic characteristics between category 4 A (low-risk group) and category 4B-5 (medium-high- risk group) MCs were identified. Finally, other ultrasonographic features of MC and FA both with an oval shape were compared to determine the key factors for differential diagnosis. The Mann-Whitney test, χ2 test or Fisher's exact test was used to compare data between groups.

RESULTS

MCs with an oval shape (81.2%) and a circumscribed margin (25%) on US were more commonly assessed in the low-risk group (BI-RADS 4 A) than in the medium-high-risk group (BI-RADS 4B-5) (20%, p < 0.001 and 0%, p = 0.001, respectively). Compared with those with FA, patients with MC were older, and tended to have masses with non-hypoechoic patterns, not circumscribed margins, and a posterior echo enhancement on US (p < 0.001, p < 0.001, and p = 0.003, respectively).

CONCLUSION

The oval shape was the main reason for the underestimation of MCs. On US, an oval mass found in the breast of women of older age with non-hypoechoic patterns, not circumscribed margins, and a posterior echo enhancement was associated with an increased risk of being an MC, and should be subjected to active biopsy.

Diagnostic Validity of Point-of-Care Breast Ultrasound for Females with Palpable Breast Masses

INTRODUCTION

With breast cancer as one of the frequent causes of cancer mortality today, the importance of ultrasound in its early detection has been apparent. It has been a valuable addition to the surgeon's diagnostic skills, contributing a vital role in clinical practice. We set out to determine the accuracy and value of breast ultrasound for primary imaging in women presenting with a clinically palpable mass in our outpatient clinic.

MATERIALS AND METHODS

This is a retrospective cross-sectional study of a point-of-care breast ultrasound among patients who consulted at the University of the Philippines-Philippine General Hospital (UP-PGH) Breast Care Clinic for a palpable breast mass without prior histopathologic diagnosis. The overall diagnostic accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were determined. Sonographic features were also identified, and multiple logistic regression analysis was performed to determine significant predictors of malignancy.

RESULTS

Eighty patients were reviewed and compared with their histopathology results. The overall accuracy of a surgeon-performed breast ultrasound was 86.2%, sensitivity of 91.4%, specificity of 82.2%, PPV of 80% and NPV of 92.5%. Indistinct borders, posterior enhancement, unilateral shadowing, heterogeneous echo pattern and deeper than wide anterior-posterior ratio are sonographic features associated with malignancy.

CONCLUSION

This study showed that a point-of-care ultrasound for a palpable breast mass is reliable with a relatively good accuracy rate. Performing breast ultrasound in the clinic will help the surgeon evaluate the extent of disease preoperatively and be guided as to the optimal surgical management for the patient.

Mixed and Purely Hyperechoic Breast Masses: A Radiologic-Pathologic Review

Breast US is a mainstay of modern-day breast imaging, especially in the diagnostic and interventional realm. The BI-RADS atlas described six echo patterns relative to the subcutaneous mammary fat: anechoic, hypoechoic, complex cystic and solid, isoechoic, heterogeneous, and hyperechoic. Hyperechoic breast masses demonstrate increased echogenicity relative to subcutaneous mammary fat or equal to fibroglandular tissue. Pathologically, the hyperechoic pattern at breast US results from the intermingling of different components: adipose tissue, fibrous tissue or stroma, secretions, blood or vascularity, and calcifications. Most hyperechoic masses are benign, especially homogeneously hyperechoic masses. However, hyperechogenicity does not exclude malignancy. Two echo patterns have been identified in hyperechoic malignant lesions, including those with a hypoechoic center and hyperechoic rim known as the rim pattern and a mass with hyperechoic areas distributed through the mass known as a dispersed pattern. This article aims to illustrate the echogenic patterns of breast lesions and various benign and malignant hyperechoic breast lesions with radiologic-pathologic correlation and to increase awareness of heterogeneously hyperechoic breast lesions as a manifestation of malignancy.

Machine learning analysis of breast ultrasound to classify triple negative and HER2+ breast cancer subtypes

OBJECTIVES

Early diagnosis of triple-negative (TN) and human epidermal growth factor receptor 2 positive (HER2+) breast cancer is important due to its increased risk of micrometastatic spread necessitating early treatment and for guiding targeted therapies. This study aimed to evaluate the diagnostic performance of machine learning (ML) classification of newly diagnosed breast masses into TN versus non-TN (NTN) and HER2+ versus HER2 negative (HER2-) breast cancer, using radiomic features extracted from grayscale ultrasound (US) b-mode images.

MATERIALS AND METHODS

A retrospective chart review identified 88 female patients who underwent diagnostic breast US imaging, had confirmation of invasive malignancy on pathology and receptor status determined on immunohistochemistry available. The patients were classified as TN, NTN, HER2+ or HER2- for ground-truth labelling. For image analysis, breast masses were manually segmented by a breast radiologist. Radiomic features were extracted per image and used for predictive modelling. Supervised ML classifiers included: logistic regression, k-nearest neighbour, and Naïve Bayes. Classification performance measures were calculated on an independent (unseen) test set. The area under the receiver operating characteristic curve (AUC), sensitivity (%), and specificity (%) were reported for each classifier.

RESULTS

The logistic regression classifier demonstrated the highest AUC: 0.824 (sensitivity: 81.8%, specificity: 74.2%) for the TN sub-group and 0.778 (sensitivity: 71.4%, specificity: 71.6%) for the HER2 sub-group.

CONCLUSION

ML classifiers demonstrate high diagnostic accuracy in classifying TN versus NTN and HER2+ versus HER2- breast cancers using US images. Identification of more aggressive breast cancer subtypes early in the diagnostic process could help achieve better prognoses by prioritizing clinical referral and prompting adequate early treatment.

Fully automatic tumor segmentation of breast ultrasound images with deep learning

BACKGROUND

Breast ultrasound (BUS) imaging is one of the most prevalent approaches for the detection of breast cancers. Tumor segmentation of BUS images can facilitate doctors in localizing tumors and is a necessary step for computer-aided diagnosis systems. While the majority of clinical BUS scans are normal ones without tumors, segmentation approaches such as U-Net often predict mass regions for these images. Such false-positive problem becomes serious if a fully automatic artificial intelligence system is used for routine screening.

METHODS

In this study, we proposed a novel model which is more suitable for routine BUS screening. The model contains a classification branch that determines whether the image is normal or with tumors, and a segmentation branch that outlines tumors. Two branches share the same encoder network. We also built a new dataset that contains 1600 BUS images from 625 patients for training and a testing dataset with 130 images from 120 patients for testing. The dataset is the largest one with pixel-wise masks manually segmented by experienced radiologists. Our code is available at https://github.com/szhangNJU/BUS_segmentation.

RESULTS

The area under the receiver operating characteristic curve (AUC) for classifying images into normal/abnormal categories was 0.991. The dice similarity coefficient (DSC) for segmentation of mass regions was 0.898, better than the state-of-the-art models. Testing on an external dataset gave a similar performance, demonstrating a good transferability of our model. Moreover, we simulated the use of the model in actual clinic practice by processing videos recorded during BUS scans; the model gave very low false-positive predictions on normal images without sacrificing sensitivities for images with tumors.

CONCLUSIONS

Our model achieved better segmentation performance than the state-of-the-art models and showed a good transferability on an external test set. The proposed deep learning architecture holds potential for use in fully automatic BUS health screening.

The tumor-infiltrating lymphocyte ultrasonography score can provide a diagnostic prediction of lymphocyte-predominant breast cancer preoperatively

PURPOSE

Tumor-infiltrating lymphocytes (TILs) are known to predict the therapeutic effect in breast cancer. Although a preoperative tissue biopsy can be used to evaluate TILs, TILs that are heterogeneously distributed might require examination of all preoperative tissue biopsy samples. We have recently reported that the TIL ultrasonography (US) score, as determined by characteristic US findings, provides excellent predictive performance for lymphocyte predominant breast cancer (LPBC). We herein aimed to determine whether the preoperative TIL-US score can more accurately predict LPBC than preoperative tissue biopsy.

METHODS

We assessed 161 patients with invasive breast cancer that were treated with curative surgery between January 2014 and December 2017. Stromal lymphocytes were examined on preoperative tissue biopsy tissues and surgical pathological specimens. Breast cancer samples with ≥ 50% stromal TILs were defined as pre-LPBC (preoperative tissue biopsy) and LPBC (surgical pathological specimens). Useful factors for predicting LPBC were searched among clinicopathological factors.

RESULTS

The TIL-US score cutoff value for predicting LPBC was 4 points based on the receiver operating characteristic curves (area under the curve: 0.88). Several significant predictors for LPBC were revealed by the undertaken multivariate logistic regression analysis (odds ratios: TIL-US score, 26.8; pre-LPBC, 18.6; HER2, 9.2; all, p < 0.05). The sensitivity, specificity, accuracy, positive predictive value, and negative predictive value were 0.74, 0.89, 0.85, 0.67, and 0.92 for the TIL-US score, respectively, and 0.51, 0.98, 0.87, 0.91, and 0.86 for the pre-LPBC, respectively.

CONCLUSION

TIL-US scores can predict LPBC preoperatively and are characterized by a significantly high sensitivity and negative predictive value.

Comparing deep learning-based automatic segmentation of breast masses to expert interobserver variability in ultrasound imaging

Deep learning is a powerful tool that became practical in 2008, harnessing the power of Graphic Processing Unites, and has developed rapidly in image, video, and natural language processing. There are ongoing developments in the application of deep learning to medical data for a variety of tasks across multiple imaging modalities. The reliability and repeatability of deep learning techniques are of utmost importance if deep learning can be considered a tool for assisting experts, including physicians, radiologists, and sonographers. Owing to the high costs of labeling data, deep learning models are often evaluated against one expert, and it is unknown if any errors fall within a clinically acceptable range. Ultrasound is a commonly used imaging modality for breast cancer screening processes and for visually estimating risk using the Breast Imaging Reporting and Data System score. This process is highly dependent on the skills and experience of the sonographers and radiologists, thereby leading to interobserver variability and interpretation. For these reasons, we propose an interobserver reliability study comparing the performance of a current top-performing deep learning segmentation model against three experts who manually segmented suspicious breast lesions in clinical ultrasound (US) images. We pretrained the model using a US thyroid segmentation dataset with 455 patients and 50,993 images, and trained the model using a US breast segmentation dataset with 733 patients and 29,884 images. We found a mean Fleiss kappa value of 0.78 for the performance of three experts in breast mass segmentation compared to a mean Fleiss kappa value of 0.79 for the performance of experts and the optimized deep learning model.

BI-RADS 3 on Screening Breast Ultrasound: What Is It and What Is the Appropriate Management?

US is widely used in breast imaging for diagnostic purposes and is also used increasingly for supplemental screening in women with dense breasts. US frequently depicts masses that are occult on mammography, even after tomosynthesis, and the vast majority of such masses are benign. Many masses seen only on screening US are easily recognized as benign simple cysts. Probably benign, BI-RADS 3, or low suspicion, BI-RADS 4A masses are also common and often prompt short-interval follow-up or biopsy, respectively, yet the vast majority of these are benign. This review details appropriate characterization, classification, and new approaches to the management of probably benign masses seen on screening US that can reduce false positives and, thereby, reduce costs and patient anxiety.

Ultrasonography Combined With Contrast-enhanced Ultrasonography Can Predict Lymphocyte-predominant Breast Cancer

BACKGROUND

We investigated whether contrast-enhanced ultrasonography (CEUS) scores can predict lymphocyte-predominant breast cancer (LPBC).

PATIENTS AND METHODS

We evaluated 75 patients who underwent US and CEUS. LPBC was defined as tissues with ≥50% stromal tumour-infiltrating lymphocytes (TILs) preoperatively. Characteristic US images predicting LPBC were evaluated using TIL-US scores via three ultrasonic tissue characteristics: Shape, internal echo level, and posterior echoes. TIL-CEUS was evaluated based on TIL-US plus CEUS.

RESULTS

TIL-US and TIL-CEUS cut-offs for predicting LPBC were 4 and 6 (area under the curve=0.93 and 0.96, respectively) points based on receiver operating characteristics curves. Sensitivity, specificity, and accuracy values (95% confidence intervaI) were 0.94 (0.77-0.99), 0.75 (0.70-0.77), and 0.80 (0.72-0.82); and 0.94 (0.78-0.99), 0.86 (0.81-0.87), and 0.88 (0.80-0.90) for TIL-US and TIL-CEUS, respectively. TIL-CEUS score was a significant single predictor for LPBC in multivariate logistic regression (p=0.001).

CONCLUSION

TIL-CEUS can be used for preoperative LPBC prediction and detection.

A new nomogram for predicting the malignant diagnosis of Breast Imaging Reporting and Data System (BI-RADS) ultrasonography category 4A lesions in women with dense breast tissue in the diagnostic setting

Background

Biopsy has been recommended for Breast Imaging Reporting and Data System (BI-RADS) category 4 lesions. However, the malignancy rate of category 4A lesions is very low (2-10%). Therefore, most biopsies of category 4A lesions are benign, and the results will generally cause additional health care costs and patient anxiety.

Methods

A prediction model was developed based on an analysis of 418 BI-RADS ultrasonography (US) category 4A patients at Sun Yat-sen Memorial Hospital. Univariate and multivariate logistic regression analyses were applied to identify significant variables for inclusion in the final nomogram. The predictive accuracy and discriminative ability were evaluated using the concordance index (C-index) and calibration curves. An independent cohort of 97 patients from the Second Affiliated Hospital of Guangzhou Medical University was used for external validation.

Results

The independent risk factors from the multivariate analysis for the training cohort were family history of breast cancer (OR =4.588, P=0.004), US features [margin (OR =2.916, P=0.019), shape (irregular vs. oval, OR =2.474, P=0.044; round vs. oval, OR =1.935, P=0.276), parallel orientation vs. not parallel (OR =2.204, P=0.040)], low suspicious lymph nodes (OR =7.664, P=0.019), and suspicious calcifications on mammography (MG) (OR =6.736, P=0.001). The C-index was good in the training [0.813, 95% confidence interval (95% CI), 0.733 to 0.893] and validation cohorts (0.765, 95% CI, 0.584 to 0.946). The calibration curves showed optimal agreement between the nomogram prediction and actual observations for the probability of malignancy. Also, the cutoff score was set to 100 for discriminating high and low risk. The model performed well in discerning different risk groups.

Conclusions

We developed a well-discriminated and calibrated nomogram to predict the malignancy of BI-RADS US category 4A lesions in dense breast tissue, which may help clinicians identify patients at lower or higher risk.

Robust Single-Probe Quantitative Ultrasonic Imaging System with a Target-Aware Deep Neural Network

OBJECTIVE

The speed of sound (SoS) has great potential as a quantitative imaging biomarker since it is sensitive to pathological changes in tissues. In this paper, a target-aware deep neural (TAD) network reconstructing an SoS image quantitatively from pulse-echo phase-shift maps gathered from a single conventional ultrasound probe is presented.

METHODS

In the proposed TAD network, the reconstruction process is guided by feature maps created from segmented target images for accuracy and contrast. In addition, the feature extraction process utilizes phase difference information instead of direct pulse-echo radio frequency (RF) data for robust image reconstruction against noise in the pulse-echo data.

RESULTS

The TAD network outperforms the fully convolutional network in root mean square error (RMSE), contrast-to-noise ratio (CNR), and structural similarity index (SSIM) in the presence of nearby reflectors. The measured RMSE and CNR are 5.4 m/s and 22 dB, respectively with the tissue attenuation coefficient of 2 dB/cm/MHz, which are 72% and 13 dB improvement over the state of the art design in RMSE and CNR, respectively. In the in-vivo test, the proposed method classifies the tissues in the neck area using SoS with a p-value below 0.025.

CONCLUSION

The proposed TAD network is the most accurate and robust single-probe SoS image reconstruction method reported to date.

SIGNIFICANCE

The accuracy and robustness demonstrated by the proposed SoS imaging method open up the possibilities of wide-spread clinical application of the single-probe SoS imaging system.

Using BI-RADS Stratifications as Auxiliary Information for Breast Masses Classification in Ultrasound Images

Breast Ultrasound (BUS) imaging has been recognized as an essential imaging modality for breast masses classification in China. Current deep learning (DL) based solutions for BUS classification seek to feed ultrasound (US) images into deep convolutional neural networks (CNNs), to learn a hierarchical combination of features for discriminating malignant and benign masses. One existing problem in current DL-based BUS classification was the lack of spatial and channel-wise features weighting, which inevitably allow interference from redundant features and low sensitivity. In this study, we aim to incorporate the instructive information provided by breast imaging reporting and data system (BI-RADS) within DL-based classification. A novel DL-based BI-RADS Vector-Attention Network (BVA Net) that trains with both texture information and decoded information from BI-RADS stratifications was proposed for the task. Three baseline models, pre-trained DenseNet-121, ResNet-50 and Residual-Attention Network (RA Net) were included for comparison. Experiments were conducted on a large scale private main dataset and two public datasets, UDIAT and BUSI. On the main dataset, BVA Net outperformed other models, in terms of AUC (area under the receiver operating curve, 0.908), ACC (accuracy, 0.865), sensitivity (0.812) and precision (0.795). BVA Net also achieved the high AUC (0.87 and 0.882) and ACC (0.859 and 0.843), on UDIAT and BUSI. Moreover, we proposed a method that integrates both BVA Net binary classification and BI-RADS stratification estimation, called integrated classification. The introduction of integrated classification helped improving the overall sensitivity while maintaining a high specificity.

Neural Network Pattern Recognition of Ultrasound Image Gray Scale Intensity Histograms of Breast Lesions to Differentiate Between Benign and Malignant Lesions: Analytical Study

Background

Ultrasound-based radiomic features to differentiate between benign and malignant breast lesions with the help of machine learning is currently being researched. The mean echogenicity ratio has been used for the diagnosis of malignant breast lesions. However, gray scale intensity histogram values as a single radiomic feature for the detection of malignant breast lesions using machine learning algorithms have not been explored yet.

Objective

This study aims to assess the utility of a simple convolutional neural network in classifying benign and malignant breast lesions using gray scale intensity values of the lesion.

Methods

An open-access online data set of 200 ultrasonogram breast lesions were collected, and regions of interest were drawn over the lesions. The gray scale intensity values of the lesions were extracted. An input file containing the values and an output file consisting of the breast lesions’ diagnoses were created. The convolutional neural network was trained using the files and tested on the whole data set.

Results

The trained convolutional neural network had an accuracy of 94.5% and a precision of 94%. The sensitivity and specificity were 94.9% and 94.1%, respectively.

Conclusions

Simple neural networks, which are cheap and easy to use, can be applied to diagnose malignant breast lesions with gray scale intensity values obtained from ultrasonogram images in low-resource settings with minimal personnel.

Bayesian Approach for a Robust Speed-of-Sound Reconstruction Using Pulse-Echo Ultrasound

Computed ultrasound tomography in echo mode (CUTE) is a promising ultrasound (US) based multi-modal technique that allows to image the spatial distribution of speed of sound (SoS) inside tissue using hand-held pulse-echo US. It is based on measuring the phase shift of echoes when detected under varying steering angles. The SoS is then reconstructed using a regularized inversion of a forward model that describes the relation between the SoS and echo phase shift. Promising results were obtained in phantoms when using a Tikhonov-type regularization of the spatial gradient (SG) of SoS. In-vivo, however, clutter and aberration lead to an increased phase noise. In many subjects, this phase noise causes strong artifacts in the SoS image when using the SG regularization. To solve this shortcoming, we propose to use a Bayesian framework for the inverse calculation, which includes a priori statistical properties of the spatial distribution of the SoS to avoid noise-related artifacts in the SoS images. In this study, the a priori model is based on segmenting the B-Mode image. We show in a simulation and phantom study that this approach leads to SoS images that are much more stable against phase noise compared to the SG regularization. In a preliminary in-vivo study, a reproducibility in the range of 10 ms^-1 was achieved when imaging the SoS of a volunteer's liver from different scanning locations. These results demonstrate the diagnostic potential of CUTE for example for the staging of fatty liver disease.

Improved forward model for quantitative pulse-echo speed-of-sound imaging

Computed ultrasound tomography in echo mode (CUTE) allows determining the spatial distribution of speed-of-sound (SoS) inside tissue using handheld pulse-echo ultrasound (US). This technique is based on measuring the changing phase of beamformed echoes obtained under varying transmit (Tx) and/or receive (Rx) steering angles. The SoS is reconstructed by inverting a forward model describing how the spatial distribution of SoS is related to the spatial distribution of the echo phase shift. Thanks to the straight-ray approximation, this forward model is linear and can be inverted in real-time when implemented in a state-of-the art system. Here we demonstrate that the forward model must contain two features that were not taken into account so far: (a) the phase shift must be detected between pairs of Tx and Rx angles that are centred around a set of common mid-angles, and (b) it must account for an additional phase shift induced by the offset of the reconstructed position of echoes. In a phantom study mimicking hepatic and cancer imaging, we show that both features are required to accurately predict echo phase shift among different phantom geometries, and that substantially improved quantitative SoS images are obtained compared to the model that has been used so far. The importance of the new model is corroborated by a preliminary volunteer result.

Harmonic motion imaging of human breast masses: an in vivo clinical feasibility

Non-invasive diagnosis of breast cancer is still challenging due to the low specificity of the imaging modalities that calls for unnecessary biopsies. The diagnostic accuracy can be improved by assessing the breast tissue mechanical properties associated with pathological changes. Harmonic motion imaging (HMI) is an elasticity imaging technique that uses acoustic radiation force to evaluate the localized mechanical properties of the underlying tissue. Herein, we studied the in vivo feasibility of a clinical HMI system to differentiate breast tumors based on their relative HMI displacements, in human subjects. We performed HMI scans in 10 female subjects with breast masses: five benign and five malignant masses. Results revealed that both benign and malignant masses were stiffer than the surrounding tissues. However, malignant tumors underwent lower mean HMI displacement (1.1 ± 0.5 µm) compared to benign tumors (3.6 ± 1.5 µm) and the adjacent non-cancerous tissue (6.4 ± 2.5 µm), which allowed to differentiate between tumor types. Additionally, the excised breast specimens of the same patients (n = 5) were imaged post-surgically, where there was an excellent agreement between the in vivo and ex vivo findings, confirmed with histology. Higher displacement contrast between cancerous and non-cancerous tissue was found ex vivo, potentially due to the lower nonlinearity in the elastic properties of ex vivo tissue. This preliminary study lays the foundation for the potential complementary application of HMI in clinical practice in conjunction with the B-mode to classify suspicious breast masses.

Tumor Proliferation and Invasiveness Derived From Ultrasound Appearances of Invasive Breast Cancers: Moving Beyond the Routine Differential Diagnosis

OBJECTIVES

To investigate the correlation between ultrasound (US) appearances of invasive breast cancers and tumor proliferation and invasiveness measured according to the histologic grade, Ki-67 expression, axillary lymph node metastasis (ALNM), and lymphovascular invasion (LVI).

METHODS

This study evaluated 676 patients who underwent primary surgical treatment of invasive breast cancers. The preoperative US reports and postoperative pathologic and immunohistochemical results of the patients were retrospectively reviewed. Ultrasound characteristics were evaluated according to the American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) lexicon. Logistic regression analyses were used to identify independent predictive US features that were correlated with tumor proliferation and invasiveness of breast cancers. Odds ratios (ORs) were calculated.

RESULTS

Posterior acoustic enhancement and calcifications on US images were independent predictive factors of a higher histologic grade and a higher Ki-67 level (OR, 1.69-6.54; P < .05). Meanwhile, a noncircumscribed margin (OR, 2.61; P < .05) and posterior acoustic shadow (OR, 1.62; P < .05) were independent predictors of ALNM. An irregular shape (OR, 2.13; P < .05) and calcifications (OR, 1.69; P < .05) were independent risk factors for LVI. Infiltrative breast cancers scored as BI-RADS category 5 had higher probability to be associated with ALNM (OR, 3.33; P < .0005) and LVI (OR, 2.87; P < .0005).

CONCLUSIONS

Ultrasound features of invasieve breast cancers might have a predictive value for tumor proliferation and invasiveness. The US features correlated with a high cellular proliferation rate were different from those associated with ALNM. The tumor shape, margin, posterior acoustic pattern, and calcifications at US are suggested to be considered by clinicians when making clinical decisions.

"The unusual suspects"-Mammographic, sonographic, and histopathologic appearance of atypical breast masses

Breast malignancy is the second most common cause of cancer death in women. However, less common breast masses can mimic carcinoma and can pose diagnostic challenges. This case-based review describes a spectrum of rare breast neoplastic and non-neoplastic masses ranging from malignant to benign entities. Malignant masses in this review include adenoid cystic carcinoma, spindle cell lipoma, granular cell tumor, angiosarcoma, glomus tumor, adenosquamous carcinoma, and myofibroblastoma. Benign masses include sarcoidosis, diabetic mastopathy, and cat scratch disease. Demographics and, when relevant, clinical presentation are summarized. Breast imaging appearance on mammography and ultrasound are highlighted along with radiology-pathology correlation with the appearance and characteristics of the histopathological specimen of these rare masses.

Pilomatrixoma, a Rare Mimicker of Male Breast Cancer

Pilomatrixoma or calcifying epithelioma of Malherbe is a benign skin tumor arising from the hair follicle; breast occurrence is considered a rarity. Clinically presenting as a palpable abnormality and with both benign and malignant mammographic and sonographic features, it can be easily misdiagnosed as a breast neoplasm. We report a very rare case of pilomatrixoma of the male breast in a 36-year-old male presenting with a firm, superficial nodule in the upper outer quadrant. Though the sonographic trifecta of imaging features (shape- margins-orientation/oval, circumscribed mass, parallel to the skin) is consistent with a benign lesion, a histologic diagnosis was warranted based on its most suspicious feature of internal pleomorphic calcifications. Pathologic diagnosis revealed the uncommon benign entity of pilomatrixoma in the male breast. Our patient was recommended for surgical excision based on current literature recommendations for management in most reports of pilomatrixoma. One alternative recommendation presented in a single report of pilomatrixoma in the breast supported follow-up imaging based on benign imaging characteristics.

Hyperechoic malignancies of the breast: Underlying pathologic features correlating with this unusual appearance on ultrasound

Hyperechogenicity in the breast on ultrasound (US) is usually regarded as a benign feature with only rare hyperechoic malignancies reported to date. In this study, we evaluated the pathologic findings on core needle biopsy of hyperechoic lesions and investigated the histologic features in malignancies that give rise to an echogenic pattern. A total of 163 core needle biopsies (CNB) were performed for "hyperechoic" or "echogenic" lesions between 1/1/05 and 7/31/17. Lesions were classified based on the proportion of hyperechoic areas identified. We found that all lesions with a homogenous hyperechoic pattern (>90% hyperechoic) were benign (n = 17), regardless of the type of margins. Malignancies were found in 21% (7/34, six invasive carcinomas and one lymphoma) of heterogenous lesions with ≥50% hyperechoic areas (all with noncircumscribed margins) and in 31% of lesions with <50% hyperechoic areas (19/61, 14 invasive carcinomas, two lymphomas, and three metastases), including five with circumscribed margins (one invasive carcinoma, one lymphoma, and three metastases). Two major US patterns were identified in malignant lesions, those with a hypoechoic center and hyperechoic rim, corresponding to a central tumor area with dense stroma and tumor cells infiltrating adipose tissue at the periphery ("rim pattern"), and a second "dispersed pattern" with hyperechoic areas distributed throughout the lesion. Hyperechoic malignancies were found to be comprised of a complex intermixture of elements of differing echogenicity including tumor cells, adipose tissue, and fluid (in tubules, stromal clefts, or blood vessels). Our findings support the importance of radiologists specifying the echogenic pattern of hyperechoic lesions, as heterogenous lesions are associated with a higher risk of malignancy and pathologists should be alert to the associated pathologic findings.

Predicting Breast Cancer in Breast Imaging Reporting and Data System (BI-RADS) Ultrasound Category 4 or 5 Lesions: A Nomogram Combining Radiomics and BI-RADS

Radiomics reflects the texture and morphological features of tumours by quantitatively analysing the grey values of medical images. We aim to develop a nomogram incorporating radiomics and the Breast Imaging Reporting and Data System (BI-RADS) for predicting breast cancer in BI-RADS ultrasound (US) category 4 or 5 lesions. From January 2017 to August 2018, a total of 315 pathologically proven breast lesions were included. Patients from the study population were divided into a training group (n = 211) and a validation group (n = 104) according to a cut-off date of March 1^st, 2018. Each lesion was assigned a category (4A, 4B, 4C or 5) according to the second edition of the American College of Radiology (ACR) BI-RADS US. A radiomics score was generated from the US image. A nomogram was developed based on the results of multivariate regression analysis from the training group. Discrimination, calibration and clinical usefulness of the nomogram for predicting breast cancer were assessed in the validation group. The radiomics score included 9 selected radiomics features. The radiomics score and BI-RADS category were independently associated with breast malignancy. The nomogram incorporating the radiomics score and BI-RADS category showed better discrimination (area under the receiver operating characteristic curve [AUC]: 0.928; 95% confidence interval [CI]: 0.876, 0.980) between malignant and benign lesions than either the radiomics score (P = 0.029) or BI-RADS category (P = 0.011). The nomogram demonstrated good calibration and clinical usefulness. In conclusion, the nomogram combining the radiomics score and BI-RADS category is potentially useful for predicting breast malignancy in BI-RADS US category 4 or 5 lesions.

Breast tumor classification using different features of quantitative ultrasound parametric images

RATIONALE AND OBJECTIVES

The ultrasound B-mode-based morphological and texture analysis and Nakagami parametric imaging have been proposed to characterize breast tumors. Since these three feature categories of ultrasonic tissue characterization supply information on different physical characteristics of breast tumors, by combining the above methods is expected to provide more clues for classifying breast tumors.

MATERIALS AND METHODS

To verify the validity of the concept, raw data were obtained from 160 clinical cases. Six different types of morphological-feature parameters, four texture features, and the Nakagami parameter of benignancy and malignancy were extracted for evaluation. The Pearson's correlation matrix was used to calculate the correlation between different feature parameters. The fuzzy c-means clustering and stepwise regression techniques were utilized to determine the optimal feature set, respectively. The logistic regression, receiver operating characteristic curve, and support vector machine were used to estimate the diagnostic ability.

RESULTS

The best performance was obtained by combining morphological-feature parameter (e.g., standard deviation of the shortest distance), texture feature (e.g., variance), and the Nakagami parameter, with an accuracy of 89.4%, a specificity of 86.3%, a sensitivity of 92.5%, and an area under receiver operating characteristic curve of 0.96. There was no significant difference between using fuzzy c-means clustering, logistic regression, and support vector machine based on the optimal feature set for breast tumors classification.

CONCLUSION

Therefore, we verified that different physical ultrasonic features are functionally complementary and thus improve the performance in diagnosing breast tumors. Moreover, the optimal feature set had the maximum discriminating performance should be irrelative to the power of classifiers.

Probably Benign Solid Breast Lesions on Ultrasound: Need for Biopsy Reassessed

Objective:

To determine the negative predictive value of ultrasound for breast masses with probably benign morphology, and to assess whether follow-up may be an acceptable alternative to biopsy.

Methods:

After Institutional Review Board approval, all solid breast masses categorized as probably benign (American College of Radiology Breast Imaging Reporting and Data System [BI-RADS] 3) on ultrasound from January 2014 to December 2015, and having either tissue diagnosis or imaging stability for 24 months, or downgrading to BIRADS 2 during imaging surveillance were included.

Result:

A total of 157 lesions in 40 patients constituted the study population. The mean patient age was 31.3 years (range, 20-56 years). Seventeen of these 157 lesions underwent tissue diagnosis with no invasive breast cancer. Out of the remaining 140 lesions, 115 were stable on imaging for 24 months or more. The rest 25 were deemed benign because of decrease in size on follow up (n=1), non-recommendation of further imaging by the second radiologist on follow up ultrasound (n= 13) or presence of benign tissue diagnosis in the largest lesion (n=11).

Conclusion:

Ultrasound has 100% negative predictive value for breast lesions with probably benign morphology, whether palpable or not. Follow up is an appropriate option to immediate biopsy of such lesions keeping in mind that noncompliance with surveillance may be a potential problem.

Masses in the era of screening tomosynthesis: Is diagnostic ultrasound sufficient?

METHODS:

All masses recalled from screening digital breast tomosynthesis between July 1, 2017 and December 31, 2017 that were sent either to diagnostic mammography or ultrasound were compared. Size, shape, margins, visibility on ultrasound, diagnostic assessment and pathology of all masses along with breast density were evaluated.

RESULTS:

102/212 digital breast tomosynthesis screen-detected masses were worked up with diagnostic mammography initially and 110/212 were worked up with ultrasound directly. There was no significant difference in ultrasound visibility of masses sent to diagnostic mammography first with those sent to ultrasound first (p = 0.42). 4 (4%) masses sent to mammogram first and 2 (2%) masses sent to ultrasound first were not visualized. There was a significant difference in size between masses that were visualized under ultrasound versus those that were not (p = 0.01), when masses in both groups were assessed cumulatively.

CONCLUSIONS:

98% of digital breast tomosynthesis screen-detected masses sent to ultrasound directly were adequately assessed without diagnostic mammography.

ADVANCES IN KNOWLEDGE:

There is potential for avoiding a diagnostic mammogram for evaluation of majority of digital breast tomosynthesis screen-detected masses.

Value of Ultrasound in the Detection of Benign and Malignant Breast Diseases: A Diagnostic Accuracy Study

Objectives

We sought to determine the diagnostic accuracy of ultrasound for benign and malignant breast lesions.

Methods

This retrospective study was performed to evaluate the diagnostic accuracy of ultrasound in 203 patients with complete medical records who visited Mehr Medical Imaging Center for breast ultrasound between March 2014 and February 2016. The collected data comprised of demographic characteristics, ultrasound results (consisting of the anatomic area of the lesion, the involved side, and the ultrasound characteristics of the lesion), mammogram results, and pathology reports (if surgery or biopsy was performed).

Results

For the diagnosis of malignant and benign lesions, ultrasound had a sensitivity of 93.9% and specificity of 86.5%; its positive and negative predictive values were 86.9% and 93.8%, respectively. Lesion type was significantly associated with a family history of breast cancer and fertility status (p < 0.005), but there was no significant association between the involved side and tumor type (p > 0.050).

Conclusions

Mammography is the best technique for screening and identifying patients with non-mass-like breast lesions and microcalcifications. Considering the false positive and false-negative results, ultrasound is not a perfect screening modality. Future studies are recommended to study the value of ultrasound in the detection of high-risk breast cancer patients.

Optoacoustic Breast Imaging: Imaging-Pathology Correlation of Optoacoustic Features in Benign and Malignant Breast Masses

OBJECTIVE

Optoacoustic ultrasound breast imaging is a fused anatomic and functional modality that shows morphologic features, as well as hemoglobin amount and relative oxygenation within and around breast masses. The purpose of this study is to investigate the positive predictive value (PPV) of optoacoustic ultrasound features in benign and malignant masses.

SUBJECTS AND METHODS

In this study, 92 masses assessed as BI-RADS category 3, 4, or 5 in 94 subjects were imaged with optoacoustic ultrasound. Each mass was scored by seven blinded independent readers according to three internal features in the tumor interior and two external features in its boundary zone and periphery. Mean and median optoacoustic ultrasound scores were compared with histologic findings for biopsied masses and nonbiopsied BI-RADS category 3 masses, which were considered benign if they were stable at 12-month follow-up. Statistical significance was analyzed using a two-sided Wilcoxon rank sum test with a 0.05 significance level.

RESULTS

Mean and median optoacoustic ultrasound scores for all individual internal and external features, as well as summed scores, were higher for malignant masses than for benign masses (p < 0.0001). High external scores, indicating increased hemoglobin and deoxygenation and abnormal vessel morphologic features in the tumor boundary zone and periphery, better distinguished benign from malignant masses than did high internal scores reflecting increased hemoglobin and deoxygenation within the tumor interior.

CONCLUSION

High optoacoustic ultrasound scores, particularly those based on external features in the boundary zone and periphery of breast masses, have high PPVs for malignancy and, conversely, low optoacoustic ultrasound scores have low PPV for malignancy. The functional component of optoacoustic ultrasound may help to overcome some of the limitations of morphologic overlap in the distinction of benign and malignant masses.

Growing BI-RADS category 3 lesions on follow-up breast ultrasound: malignancy rates and worrisome features

OBJECTIVE

  To determine the frequency and malignancy rate among growing Breast Imaging Reporting and Data System (BI-RADS) category 3 lesions diagnosed on follow-up breast ultrasound and to evaluate the radiological and clinical features associated with malignancy.  Methods:  Of the 38,482 women who underwent breast ultrasound between January 2010 and December 2011, 11,582 (30.1%) patients had 12,514 BI-RADS category 3 lesions. Patients whose lesions showed ≥20% enlargement on follow-up ultrasound were selected for this study. Radiological and clinical features including increase in the maximum diameter and anteroposterior dimension, morphological changes determined via ultrasound, palpability, multiplicity, new mass, baseline breast ultrasound indication and mammographic BI-RADS category were evaluated to determine their association with malignancy. Multivariate analyses were used to identify independent predictors of malignancy.  Results:  The frequency of growing BI-RADS category 3 lesions on follow-up ultrasound was 5.9% (738 of 12,514). Of 527 lesions examined in 459 patients with a follow-up duration of at least 24 months or with available pathological results, 26 proved to be malignant (4.9%). Multivariate analyses further indicated that sonographic morphological changes (OR, 7.662, p < 0.001) and development of suspicious features on follow-up mammography (OR, 3.812, p = 0.009) were associated with malignancy. Enlargement without associated suspicious mammography or sonographic morphological abnormalities had only 1.9 % (BI-RADS category 3) chance of malignancy.  Conclusion:  The malignancy rate for growing BI-RADS category 3 lesions is 4.9%. Sonographic morphological changes and suspicious mammographic features in these tumors are significantly associated with malignancy. Advances in knowledge: For lesions with an interval growth in the anteroposterior dimension of ≤50% without morphological changes, together with a benign mammogram, follow-up rather than an immediate biopsy can be recommended to reduce false-positive biopsy results. The risk of malignancy in lesions with a size increment but with no morphologic change on sonography is only 1.9%, compatible with continued BI-RADS category 3 classification.

Diagnosing Breast Cancer with Microwave Technology: remaining challenges and potential solutions with machine learning

Currently, breast cancer often requires invasive biopsies for diagnosis, motivating researchers to design and develop non-invasive and automated diagnosis systems. Recent microwave breast imaging studies have shown how backscattered signals carry relevant information about the shape of a tumour, and tumour shape is often used with current imaging modalities to assess malignancy. This paper presents a comprehensive analysis of microwave breast diagnosis systems which use machine learning to learn characteristics of benign and malignant tumours. The state-of-the-art, the main challenges still to overcome and potential solutions are outlined. Specifically, this work investigates the benefit of signal pre-processing on diagnostic performance, and proposes a new set of extracted features that capture the tumour shape information embedded in a signal. This work also investigates if a relationship exists between the antenna topology in a microwave system and diagnostic performance. Finally, a careful machine learning validation methodology is implemented to guarantee the robustness of the results and the accuracy of performance evaluation.

B-mode ultrasound examination of canine mammary gland neoplastic lesions of small size (diameter < 2 cm)

Ultrasonography is a valuable tool for the evaluation of neoplastic lesions in the dog and there is a growing interest in the use of this technique for the stadiation of canine mammary tumours. An accurate assessment of small sized nodules facilitates the stadiation of the mammary lesions and helps the clinician in the choice of the most indicated surgical therapy. The aim of this study was to identify those ultrasound criteria that may be useful in discriminating between benign and malignant lesions of small size (diameter smaller than 2 cm). Sixty-two nodules, < 2 cm in larger diameter, belonging to thirty-five bitches presented between January 2012 and February 2014 were evaluated. Tumours were observed by conventional ultrasound and assessed for: shape (regular-irregular), limit (defined-ill-defined), margins (regular-irregular), echogenicity (hypoechoic-isoechoic-hyperecoic), echotexture (homogeneus-heterogeneus), presence of hyperecoic halo, distal acoustic enhancement or shadowing and surrounding tissue alterations. Among the alterations in surrounding tissues, the disruption of the glandular tissue and the increase in echogenicity of the peritumoral tissues were assessed. Thereafter, bitches were subjected to mastectomy and nodules were evaluated histologically. None of the ultasound criteria considered in the current study showed a statistically significant relation with malignancy, except for the presence of alterations in the tissue surrounding the nodules. According to our results, this characteristic may indicate malignancy, however its subjectivity may affect the applicability in clinical practice. In conclusions, conventional ultrasound in bitches had a limited ability in discriminating benign and malignant mammary gland neoplastic lesions of small size (diameter < 2 cm).

Application of FT-IR spectroscopy on breast cancer serum analysis

Breast cancer is regarded as the most malignant tumor among women throughout the world. Therefore, early detection and proper diagnostic methods have been known to help save women's lives. Fourier Transform Infrared (FT-IR) spectroscopy, coupled with PCA-LDA analysis, is a new technique to investigate the characteristics of serum in breast cancer. In this study, 43 breast cancer and 43 healthy serum samples were collected, and the FT-IR spectra were recorded for each one. Then, PCA analysis and linear discriminant analysis (LDA) were used to analyze the spectral data. The results showed that there were differences between the spectra of the two groups. Discriminating wavenumbers were associated with several spectral differences over the 950-1200cm^-1(sugar), 1190-1350cm^-1 (collagen), 1475-1710cm^-1 (protein), 1710-1760cm^-1 (ester), 2800-3000cm^-1 (stretching motions of -CH₂ & -CH₃), and 3090-3700cm^-1 (NH stretching) regions. PCA-LDA performance on serum IR could recognize changes between the control and the breast cancer cases. The diagnostic accuracy, sensitivity, and specificity of PCA-LDA analysis for 3000-3600cm^-1 (NH stretching) were found to be 83%, 84%, 74% for the control and 80%, 76%, 72% for the breast cancer cases, respectively. The results showed that the major spectral differences between the two groups were related to the differences in protein conformation in serum samples. It can be concluded that FT-IR spectroscopy, together with multivariate data analysis, is able to discriminate between breast cancer and healthy serum samples.

Sparse Representation Based Multi-Instance Learning for Breast Ultrasound Image Classification

We propose a novel method based on sparse representation for breast ultrasound image classification under the framework of multi-instance learning (MIL). After image enhancement and segmentation, concentric circle is used to extract the global and local features for improving the accuracy in diagnosis and prediction. The classification problem of ultrasound image is converted to sparse representation based MIL problem. Each instance of a bag is represented as a sparse linear combination of all basis vectors in the dictionary, and then the bag is represented by one feature vector which is obtained via sparse representations of all instances within the bag. The sparse and MIL problem is further converted to a conventional learning problem that is solved by relevance vector machine (RVM). Results of single classifiers are combined to be used for classification. Experimental results on the breast cancer datasets demonstrate the superiority of the proposed method in terms of classification accuracy as compared with state-of-the-art MIL methods.

Clinical significance of interval changes in breast lesions initially categorized as probably benign on breast ultrasound

The aims of this study were to determine the malignancy rate of probably benign lesions that show an interval change on follow-up ultrasound and to evaluate the differences seen on imaging between benign and malignant lesions initially categorized as probably benign but with interval change on follow-up breast ultrasound.We retrospectively reviewed 11,323 lesions from ultrasound-guided core-biopsies performed between June 2004 and December 2014 and identified 289 lesions (266 patients) with an interval change from probably benign (Breast Imaging Reporting and Data System [BI-RADS] category 3) in the previous 2 years. Malignancy rates were compared according to the ultrasound findings and the characteristics of the interval changes, including changes in morphology and/or diameter.The malignancy rate for probably benign lesions that showed an interval change on follow-up ultrasound was 6.9% (20/289). The malignancy rate was higher for clustered cysts (33.3%) and irregular or noncircumscribed masses (12.7%) than for circumscribed oval masses (5%) or complicated cysts (5%) seen on initial ultrasound (P = 0.043). Fifty-five percent of the malignancies were found to be ductal carcinoma in situ and there was 1 case of lymph node metastasis among the patients with invasive disease in whom biopsy was delayed by 6 to 15 months. The extent of invasiveness was greater in missed cases. There was a significant difference in the maximal diameter change between the 20 malignant lesions and the 269 benign lesions (4.0 mm vs 2.7 mm, P = 0.002). The cutoff value for maximal diameter change per initial diameter was 39.0% for predicting malignancy (sensitivity 95%, specificity 53.5%). The malignancy rate for morphologically changed lesions was significantly higher than for morphologically stable lesions (13.6% vs 4.9%; P = 0.024)Our 6.9% of probably benign lesions that showed an interval change finally turned out to be malignancy was mostly DCIS. The sonographic features, interval changes in sonographic features, and lesion size might help in the recategorization of these lesions.

Ultrasonography and histology correlation in BI-RADS 4/5 small breast lesions among Thai patients

Background: Ultrasonography is an important imaging tool in detection of small breast cancers, particularly in younger women with dense breasts. Among the ultrasonographic characteristics for the malignancy, it is unclear which are common or more predictive.

Objective: Analyze breast ultrasonograms and determine the common and predictive characteristics of the BIRADS 4/5 small breast lesions that were correlated with histology-proved carcinoma among Thai patients.

Methods: Data were collected retrospectively between November 2006 and September 2007 at King Chulalongkorn Memorial Hospital. Forty-five BI-RADS 4 or 5 small breast lesions from 41 patients were reviewed for ultrasonographic characteristics and for correlation between each of these features and histology-proved malignancy.

Results: There were 15 out of 30 lesions of BI-RADS 4 and 14 out of 15 of BI-RADS 5 that were histologically proven breast carcinoma. The lesion dimension ranged from 0.27 cm to 1.5 cm (mean: 0.98 cm). The malignant signs that were common consisted of irregular shapes (70%) and posterior shadowing (35.6%). However, the most correlating signs for malignancy were vascularity of the lesion 100%, and spiculated margins 100%. The other characteristics for malignancy, in descending order, were marked hypoechoicity 88.9%, microcalcifications within mass 85.7%, echogenic halo 83.3%, shadowing 81.3%, branched pattern 77.8%, duct extension 75%, irregular shape72.2%, and taller than wide orientation 70%.

Conclusion: Irregular shape and shadowing were the two most common malignant signs that characterized BI-RADS 4, 5 small breast lesions by ultrasonography. However, the most predictive signs were increases in vascularity and spiculated margins.

Diagnostic Accuracy of Contrast-Enhanced Ultrasound Enhancement Patterns for Thyroid Nodules

Background

The aim of this study was to investigate the accuracy of contrast-enhanced ultrasound (CEUS) enhancement patterns in the assessment of thyroid nodules.

Material/Methods

A total of 158 patients with suspected thyroid cancer underwent conventional ultrasound (US) and CEUS examinations. The contrast enhancement patterns of the lesions, including the peripheries of the lesions, were assessed by CEUS scans. The relationship between the size of the lesions and the degree of enhancement was also studied. US- and/or CEUS-guided biopsy was used to obtain specimens for histopathological diagnosis.

Results

The final data included 148 patients with 157 lesions. Seventy-five patients had 82 malignant lesions and 73 patients had 75 benign lesions. Peripheral ring enhancement was seen in 40 lesions. The differences of enhancement patterns and peripheral rings between benign and malignant nodules were significant (p=0.000, 0.000). The diagnostic sensitivity, specificity, and accuracy for malignant were 88%, 65.33%, and 88.32%, respectively, for CEUS, whereas they were 98.33%, 42.67%, and 71.97%, respectively, for TC by conventional US. The misdiagnosis rate by conventional US was 57.33% and 34.67% by CEUS (p=0.005). With regard to the size of lesions, a significant difference was found between low-enhancement, iso-enhancement, high-enhancement, iso-enhancement with no-enhancement area and no-enhancement (p=0.000).

Conclusions

In patients with suspicious US characteristics, CEUS had high specificity and contributed to establishing the diagnosis. Therefore, CEUS could avoid unnecessary biopsy.

Value of Virtual Touch Tissue Imaging Quantification for Evaluation of Ultrasound Breast Imaging-Reporting and Data System Category 4 Lesions

The purpose of the study was to evaluate the value of 2-D shear wave elastography (SWE) of virtual touch tissue imaging quantification (VTIQ) for ultrasound (US) Breast Imaging-Reporting and Data System (BI-RADS) category 4 lesions. One hundred sixteen lesions were subject to conventional US, conventional strain elastography (SE) of elasticity imaging (EI), acoustic radiation force impulse (ARFI)-induced SE of virtual touch tissue imaging (VTI) and VTIQ before biopsies. Of the 116 lesions, 69 (59.5%) were benign and 47 (40.5%) were malignant. Significant differences were found between benign and malignant lesions in EI score, VTI score and shear wave speed (SWS) on VTIQ (both p < 0.05). The cut-off values were EI score ≥4, VTI score ≥4 and SWS ≥3.49 m/s, respectively. The diagnostic performance of VTIQ in terms of area under receiver operating characteristic curve (AUROC) were the highest (i.e., AUROC = 0.907), in comparison with EI, VTI alone or a combination of both. The associated sensitivity, specificity and accuracy were 87.2%, 82.6% and 84.5%, respectively. The combination of VTI and VTIQ, however, was similar with US BI-RADS (p = 0.475) in sensitivity in that only two (4.3%) of 47 malignant lesions were misdiagnosed as benign that were BI-RADS category 4b on US. VTIQ is valuable to differentiate benign from malignant BI-RADS category 4 lesions, and the combination of VTI and VTIQ might be useful for patient selection before biopsy.

Polyvinyl chloride plastisol breast phantoms for ultrasound imaging

Ultrasonic phantoms are objects that mimic some features of biological tissues, allowing the study of their interactions with ultrasound (US). In the diagnostic-imaging field, breast phantoms are an important tool for testing performance and optimizing US systems, as well as for training medical professionals. This paper describes the design and manufacture of breast lesions by using polyvinyl chloride plastisol (PVCP) as the base material. Among the materials available for this study, PVCP was shown to be stable, durable, and easy to handle. Furthermore, it is a nontoxic, nonpolluting, and low-cost material. The breast's glandular tissue (image background) was simulated by adding graphite powder with a concentration of 1% to the base material. Mixing PVCP and graphite powder in differing concentrations allows one to simulate lesions with different echogenicity patterns (anechoic, hypoechoic, and hyperechoic). From this mixture, phantom materials were obtained with speed of sound varying from 1379.3 to 1397.9ms(-1) and an attenuation coefficient having values between 0.29 and 0.94dBcm(-1) for a frequency of 1MHz at 24°C. A single layer of carnauba wax was added to the lesion surface in order to evaluate its applicability for imaging. The images of the phantoms were acquired using commercial ultrasound equipment; a specialist rated the images, elaborating diagnoses representative of both benign and malignant lesions. The results indicated that it was possible to easily create a phantom by using low-cost materials, readily available in the market and stable at room temperature, as the basis of ultrasonic phantoms that reproduce the image characteristics of fatty breast tissue and typical lesions of the breast.

Predictive performance of BI-RADS magnetic resonance imaging descriptors in the context of suspicious (category 4) findings

Objective

To determine the positive predictive value (PPV) and likelihood ratio for magnetic resonance imaging (MRI) characteristics of category 4 lesions, as described in the Breast Imaging Reporting and Data System (BI-RADS^®) lexicon, as well as to test the predictive performance of the descriptors using multivariate analysis and the area under the curve derived from a receiver operating characteristic (ROC) curve.

Materials and Methods

This was a double-blind review study of 121 suspicious findings from 98 women examined between 2009 and 2013. The terminology was based on the 2013 edition of the BI-RADS.

Results

Of the 121 suspicious findings, 53 (43.8%) were proven to be malignant lesions, with no significant difference between mass and non-mass enhancement (p = 0.846). The PPVs were highest for masses with a spiculated margin (71%) and round shape (63%), whereas segmental distribution achieved a high PPV (80%) for non-mass enhancement. Kinetic analyses performed poorly, except for type 3 curves applied to masses (PPV of 73%). Logistic regression models were significant for both patterns, although the results were better for masses, particularly when kinetic assessments were included (p = 0.015; pseudo R² = 0.48; area under the curve = 90%).

Conclusion

Some BI-RADS MRI descriptors have high PPV and good predictive performance-as demonstrated by ROC curve and multivariate analysis-when applied to BI-RADS category 4 findings. This may allow future stratification of this category.

New Fully Automated Method for Segmentation of Breast Lesions on Ultrasound Based on Texture Analysis

The study described here explored a fully automatic segmentation approach based on texture analysis for breast lesions on ultrasound images. The proposed method involves two main stages: (i) In lesion region detection, the original gray-scale image is transformed into a texture domain based on log-Gabor filters. Local texture patterns are then extracted from overlapping lattices that are further classified by a linear discriminant analysis classifier to distinguish between the "normal tissue" and "breast lesion" classes. Next, an incremental method based on the average radial derivative function reveals the region with the highest probability of being a lesion. (ii) In lesion delineation, using the detected region and the pre-processed ultrasound image, an iterative thresholding procedure based on the average radial derivative function is performed to determine the final lesion contour. The experiments are carried out on a data set of 544 breast ultrasound images (including cysts, benign solid masses and malignant lesions) acquired with three distinct ultrasound machines. In terms of the area under the receiver operating characteristic curve, the one-way analysis of variance test (α=0.05) indicates that the proposed approach significantly outperforms two published fully automatic methods (p<0.001), for which the areas under the curve are 0.91, 0.82 and 0.63, respectively. Hence, these results suggest that the log-Gabor domain improves the discrimination power of texture features to accurately segment breast lesions. In addition, the proposed approach can potentially be used for automated computer diagnosis purposes to assist physicians in detection and classification of breast masses.

Validation of the fifth edition BI-RADS ultrasound lexicon with comparison of fourth and fifth edition diagnostic performance using video clips

Purpose

The aim of this study was to evaluate the positive predictive value (PPV) and the diagnostic performance of the ultrasonographic descriptors in the fifth edition of BI-RADS, comparing with the fourth edition using video clips.

Methods

From September 2013 to July 2014, 80 breast masses in 74 women (mean age, 47.5±10.7 years) from five institutions of the Korean Society of Breast Imaging were included. Two radiologists individually reviewed the static and video images and analyzed the images according to the fourth and fifth edition of BI-RADS. The PPV of each descriptor was calculated and diagnostic performances between the fourth and fifth editions were compared.

Results

Of the 80 breast masses, 51 (63.8%) were benign and 29 (36.2%) were malignant. Suspicious ultrasonographic features such as irregular shape, non-parallel orientation, angular or spiculated margins, and combined posterior features showed higher PPV in both editions (all P<0.05). No significant differences were found in the diagnostic performances between the two editions (all P>0.05). The area under the receiver operating characteristics curve was higher in the fourth edition (0.708 to 0.690), without significance (P=0.416).

Conclusion

The fifth edition of the BI-RADS ultrasound lexicon showed comparable performance to the fourth edition and can be useful in the differential diagnosis of breast masses using ultrasonography.

Evaluation of Malignancy Grade of Breast Cancer Using Perflubutane-Enhanced Ultrasonography

Whether the contrast effects of perflubutane on contrast-enhanced ultrasonography can predict the malignancy grade of breast cancer is unknown. We analyzed associations between perfusion parameters created from time-intensity curves based on enhancement intensity and temporal changes in contrast-enhanced ultrasonography and clinicopathologic factors in 100 consecutive patients with invasive breast cancer. Values of perfusion parameters were significantly greater in estrogen receptor-negative than -positive tumors (peak intensity, p = 0.0002; ascending slope, p = 0.006; area under the curve, p = 0.0006). Variations in the peak intensity of Ki-67 were significantly correlated in all tumors (r = 0.54, p < 0.0001) and in luminal (r = 0.43, p = 0.0002), human epidermal growth factor receptor type 2-positive (r = 0.47, p = 0.047) and triple-negative (r = 0.55, p = 0.043) tumors. Perfusion parameters on contrast-enhanced ultrasonography can provide excellent predictive value for high-grade malignancy and might help to determine appropriate therapeutic strategies.

Predictors of Malignancy in Hyperechoic Breast Lesions

OBJECTIVES

Hyperechogenicity has been strongly associated with benign breast lesions. Although it is correct in most cases, hyperechogenicity must not always be considered synonymous with benignancy, as hyperechoic breast cancers do occur. The purpose of this study was to review clinical and imaging characteristics of hyperechoic breast lesions, looking for features associated with malignancy.

METHODS

Institutional Review Board approval was granted for this research. A total of 19,417 sonographic examinations were performed between January 2009 and June 2013. Among these, hyperechoic lesions with histologic diagnoses, stability on long-term followup, or characteristic imaging appearances were included in the study. The patients' clinical charts, mammograms, and sonograms were reviewed. The clinical and imaging features were recorded, and the data was analyzed by the χ(2) test, Fisher exact test, and independent-samples t test, looking for statistically significant predictors of malignancy.

RESULTS

Among the 19,417 scans, 42 patients (0.2%) with 44 hyperechoic lesions were identified. Twenty-six lesions fulfilling the inclusion criteria were included in the study: 5 malignancies (3 invasive ductal carcinomas, 1 invasive lobular carcinoma, and 1 invasive mucinous cancer) and 21 benign lesions. An irregular shape, a nonparallel orientation, and noncircumscribed margins were significantly associated with the risk of malignancy (P = .002, .02, and .01, respectively).

CONCLUSIONS

A hyperechoic breast lesion must not always be assumed to be benign. Instead, a full sonographic assessment according to the American College of Radiology Breast Imaging Reporting and Data System descriptors is needed for correct characterization and avoidance of misdiagnosis.

Variability in Observer Performance Between Faculty Members and Residents Using Breast Imaging Reporting and Data System (BI-RADS)-Ultrasound, Fifth Edition (2013)

Background

Ultrasonography (US) is a useful tool for breast imaging, yet is highly operator-dependent.

Objectives

We evaluated inter-observer variability and performance discrepancies between faculty members and radiology residents when describing breast lesions, by the fifth edition of breast imaging reporting and data system (BI-RADS)-US lexicon, and then attempted to identify whether inter-observer variability could be improved after one education session.

Patients and Methods

In total, 50 malignant lesions and 70 benign lesions were considered in our retrospective study. Two faculty members, two senior residents, and two junior residents separately assessed the US images. After the first assessment, the readers received one education session, and then reassessed the images in a random order. Inter-observer variability was measured using the kappa coefficient (κ). Performance discrepancy was evaluated by receiver operating characteristic (ROC) curves.

Results

For the faculty members, fair-to-good agreement was obtained in all descriptors and final assessment, while for residents, poor-to-moderate agreement was obtained. The areas under the ROC curves were 0.78 for the faculty members, 0.59 for the senior residents, and 0.52 for the junior residents, respectively. Diagnostic performance was significantly higher in the faculty members than the senior and junior residents (P = 0.0001 and < 0.0001, respectively). After one education session, the agreement in the final assessment was one level higher in the faculty members and senior residents, yet in the senior residents, the degree of agreement was still only fair. Moreover, in the junior residents, there was no improvement.

Conclusion

Investigative assessment of breast US by residents is inadvisable. We recommend continued professional resident training to improve the degree of agreement and performance.