Purpose: The goal of this study was to follow benign thyroid nodules confirmed by ultrasound (US)-guided core needle biopsy (CNB) after inconclusive cytology on fine-needle aspiration (FNA) biopsy. Methods: Sixty-two thyroid nodules from 62 patients with CNB-confirmed benign histology that initially had inconclusive cytology on FNA were retrospectively included. The thyroid nodules were followed for 38.7 months (median, 27.5 months; range, 6 to 101 months), and the US findings of biopsied nodules, such as the interval change in size, US characteristics, and imaging category based on the Korean Thyroid Imaging Reporting and Data System (K-TIRADS), were evaluated. In addition, patients’ clinical records were reviewed for any further management or newly diagnosed thyroid malignancy. Results: Among 62 cases, three (4.8%) showed interval size growth, while 59 (95.2%) demonstrated no interval change or a decrease in size. There was no upgrade of K-TIRADS category or any newly diagnosed malignancy during the follow-up period. Conclusion US-guided CNB-confirmed benign thyroid nodules with inconclusive cytology on FNA showed a stable status during follow-up, and repeated CNB could be helpful in the management of nodules with inconclusive cytology on FNA.

Purpose: To evaluate a deep learning model to predict recurrence of thyroid tumor using preoperative ultrasonography (US). Materials and Methods: We included representative images from 229 US-based patients (male:female = 42:187; mean age, 49.6 years) who had been diagnosed with thyroid cancer on preoperative US and subsequently underwent thyroid surgery. After selecting each representative transverse or longitudinal US image, we created a data set from the resulting database of 898 images after augmentation. The Python 2.7.6 and Keras 2.1.5 framework for neural networks were used for deep learning with a convolutional neural network. We compared the clinical and histological features between patients with and without recurrence. The predictive performance of the deep learning model between groups was evaluated using receiver operating characteristic (ROC) analysis, and the area under the ROC curve served as a summary of the prognostic performance of the deep learning model to predict recurrent thyroid cancer. Results: Tumor recurrence was noted in 49 (21.4%) among the 229 patients. Tumor size and multifocality varied significantly between the groups with and without recurrence (p < 0.05). The overall mean area under the curve (AUC) value of the deep learning model for prediction of recurrent thyroid cancer was 0.9 ± 0.06. The mean AUC value was 0.87 ± 0.03 in macrocarcinoma and 0.79 ± 0.16 in microcarcinoma. Conclusion A deep learning model for analysis of US images of thyroid cancer showed the possibility of predicting recurrence of thyroid cancer.

PURPOSE: To investigate whether a correlation exists between the semi-quantitative breast-specific gamma imaging (BSGI) findings and dynamic contrast-enhanced (DCE) MRI parameters assessed by a computer-aided evaluation program. MATERIALS AND METHODS: Semi-quantitative index of the lesion to non-lesion (L/N) ratio in BSGI and DCE-MRI parameters was assessed by a computer-aided evaluation program, where 47 cases of invasive breast cancer were obtained. Correlations between the L/N ratio and DCE-MRI parameters were assessed by a computer-aided evaluation program. Tumor diameter (cm), angio-volume (cc), degree of initial peak enhancement (%), persistent enhancement proportion (%), and washout enhancement proportion (%) were analysed. The relationships between the L/N ratio and DCE-MRI parameters were evaluated by a univariate and multivariate regression analysis. RESULTS: The mean L/N ratio of the 47 tumors was 3.63 ± 2.19 (range: 1–13.1). The L/N ratio was higher in tumors with larger diameters (p < 0.001), increased angio-volume (p < 0.001), higher degree of initial peak enhancement (p = 0.005) and increased washout enhancement proportion (p = 0.004). In the multivariate regression analysis, angio-volume (cc) and washout enhancement proportion (%) were associated with L/N ratio (p = 0.007 and p = 0.024, respectively). CONCLUSION There was a correlation between the semi-quantitative L/N ratio in BSGI and DCE-MRI parameters assessed by a computer-aided evaluation program for breast cancer.

OBJECTIVE: To evaluate the additional effect of sonoelastography on the radiologist's ability for distinguishing benign from malignant complex breast masses and to decide whether to perform biopsy by B-mode US. MATERIALS AND METHODS: One hundred eighteen complex breast masses (15 malignant lesions, 103 benign lesions) were included. Five blinded readers independently assessed the likelihood of the malignancy score from 1 to 5 for two data sets (B-mode ultrasound alone and B-mode ultrasound with sonoelastography). Elasticity scores were categorized as 0, 1, or 2 based on the degree and distribution of strain of the echogenic component within complex masses. The readers were asked to downgrade the likelihood of the malignancy score when an elasticity score of 0 was assigned and to upgrade the likelihood of the malignancy score when an elasticity score of 2 was assigned. The likelihood of the malignancy score was maintained as it was for the lesions with an elasticity score of 1. The Az values, sensitivities, and specificities were compared. RESULTS: The Az value of B-mode ultrasound with sonoelastography (mean, 0.863) was greater than that of B-mode ultrasound alone (mean, 0.731; p = 0.001-0.007) for all authors. The specificity of B-mode ultrasound with sonoelastography (mean, 37.1%) was greater than that of B-mode ultrasound alone (mean, 3.8%; p < 0.001) for all readers. The addition of sonoelastography led to changes in decisions. A mean of 33.6% of benign masses were recommended for follow-up instead of biopsy. CONCLUSION: For complex breast masses, sonoelastography allows increase in both the accuracy in distinguishing benign from malignant lesions and the specificity in deciding whether to perform biopsy.
Adult ; Aged ; Biopsy/*methods ; Breast Diseases/pathology/*ultrasonography ; Breast Neoplasms/diagnosis ; Diagnosis, Differential ; Elasticity Imaging Techniques/*methods ; Female ; Humans ; Middle Aged ; Pregnancy ; Reproducibility of Results ; Retrospective Studies ; *Ultrasonography, Mammary ; Young Adult

PURPOSE: To report on the current practices in breast magnetic resonance imaging (MRI) in Korea. MATERIALS AND METHODS: We invited the 68 members of the Korean Society of Breast Imaging who were working in hospitals with available breast MRI to participate in a survey on how they performed and interpreted breast MRI. We asked one member from each hospital to respond to the survey. A total of 22 surveys from 22 hospitals were analyzed. RESULTS: Out of 22 hospitals, 13 (59.1%) performed at least 300 breast MRI examinations per year, and 5 out of 22 (22.7%) performed > 1200 per year. Out of 31 machines, 14 (45.2%) machines were 1.5-T scanners and 17 (54.8%) were 3.0-T scanners. All hospitals did contrast-enhanced breast MRI. Full-time breast radiologists supervised the performance and interpreted breast MRI in 19 of 22 (86.4%) of hospitals. All hospitals used BI-RADS for MRI interpretation. For computer-aided detection (CAD), 13 (59.1%) hospitals sometimes or always use it and 9 (40.9%) hospitals did not use CAD. Two (9.1%) and twelve (54.5%) hospitals never and rarely interpreted breast MRI without correlating the mammography or ultrasound, respectively. The majority of respondents rarely (13/21, 61.9%) or never (5/21, 23.8%) interpreted breast MRI performed at an outside facility. Of the hospitals performing contrast-enhanced examinations, 15 of 22 (68.2%) did not perform MRI-guided interventional procedures. CONCLUSION: Breast MRI is extensively performed in Korea. The indication and practical patterns are diverse. The information from this survey would provide the basis for the development of Korean breast MRI practice guidelines.
Breast Neoplasms ; Breast* ; Diagnosis ; Korea ; Magnetic Resonance Imaging* ; Mammography ; Surveys and Questionnaires ; Ultrasonography

OBJECTIVE: To retrospectively evaluate the features of undiagnosed breast cancers on prior screening breast magnetic resonance (MR) images in patients who were subsequently diagnosed with breast cancer, as well as the potential utility of MR-computer-aided evaluation (CAE). MATERIALS AND METHODS: Between March 2004 and May 2013, of the 72 consecutive pairs of prior negative MR images and subsequent MR images with diagnosed cancers (median interval, 32.8 months; range, 5.4-104.6 months), 36 (50%) had visible findings (mean size, 1.0 cm; range, 0.3-5.2 cm). The visible findings were divided into either actionable or underthreshold groups by the blinded review by 5 radiologists. MR imaging features, reasons for missed cancer, and MR-CAE features according to actionability were evaluated. RESULTS: Of the 36 visible findings on prior MR images, 33.3% (12 of 36) of the lesions were determined to be actionable and 66.7% (24 of 36) were underthreshold; 85.7% (6 of 7) of masses and 31.6% (6 of 19) of non-mass enhancements were classified as actionable lesions. Mimicking physiologic enhancements (27.8%, 10 of 36) and small lesion size (27.8%, 10 of 36) were the most common reasons for missed cancer. Actionable findings tended to show more washout or plateau kinetic patterns on MR-CAE than underthreshold findings, as the 100% of actionable findings and 46.7% of underthreshold findings showed washout or plateau (p = 0.008). CONCLUSION: MR-CAE has the potential for reducing the number of undiagnosed breast cancers on screening breast MR images, the majority of which are caused by mimicking physiologic enhancements or small lesion size.
Adult ; Aged ; Breast/*pathology ; Breast Neoplasms/*diagnosis/pathology/radiography ; Diagnosis, Computer-Assisted/*methods ; False Negative Reactions ; Female ; Humans ; Magnetic Resonance Imaging/*methods ; Mammography/*methods ; Middle Aged ; Retrospective Studies

Immunoglobulin G4 (IgG4)-related disease is a systemic fibro-inflammatory disease characterized by pathologic findings in various organs. Imaging is critical for the diagnosis and treatment assessment of patients with IgG4-related disease. In this pictorial essay, we review the key features of multiple imaging modalities, typical pathologic findings, and differential diagnosis of IgG4-related disease. This systematic pictorial review can further our understanding of the broad-spectrum manifestations of this disease.

Immunoglobulin G4 (IgG4)-related disease is a systemic fibro-inflammatory disease characterized by pathologic findings in various organs. Imaging is critical for the diagnosis and treatment assessment of patients with IgG4-related disease. In this pictorial essay, we review the key features of multiple imaging modalities, typical pathologic findings, and differential diagnosis of IgG4-related disease. This systematic pictorial review can further our understanding of the broad-spectrum manifestations of this disease.

PURPOSE: The purpose of this study was to assess the value of adding digital breast tomosynthesis (DBT) to full-field digital mammography (FFDM) in the diagnostic workup of breast cancer and to determine which lesion variables affect cancer detectability in the combined modality. METHODS: Between March and May 2012, paired FFDM and DBT images were obtained from 203 women as part of a diagnostic workup for breast cancer. Images from FFDM alone, DBT alone, and DBT combined with FFDM were reviewed in separate sessions by six blinded readers. Jackknife alternative free-response receiver operating characteristic (JAFROC) figure of merit (FOM), sensitivity, and specificity were compared between the modalities. Lesion characteristics affecting the cancer detection rate when using the combined modality were also analyzed. RESULTS: Among the 203 women, 126 women had a total of 129 malignancies and 77 women had total of 77 benign lesions. The overall JAFROC FOM of the combined modality was higher than that of FFDM alone (0.827 vs. 0.775, p<0.001) and that of DBT alone was higher than that of FFDM alone (0.807 vs. 0.775, p=0.027). The overall sensitivity of the combined modality was higher than that of FFDM alone (80.0% vs. 73.2%, p<0.001) and that of DBT alone was higher than that of FFDM alone (78.3% vs. 73.2%, p=0.007). Compared to FFDM alone, the combined modality detected an additional 48 cancers. Using the combined modality, the presence of masses or microcalcifications was significantly associated with the cancer detection rate (p<0.001). CONCLUSION: The combination of DBT with FFDM results in a higher diagnostic yield than FFDM alone. Additionally, DBT alone performs better than FFDM alone. However, even when DBT is combined with FFDM, breast cancers with no discernible masses and those lacking calcifications are difficult to detect.
Breast Neoplasms ; Breast* ; Early Detection of Cancer ; Female ; Humans ; Imaging, Three-Dimensional ; Mammography* ; ROC Curve ; Sensitivity and Specificity