3D Breast Cancer Segmentation in DCE-MRI Using Deep Learning With Weak Annotation.

BACKGROUND: Deep learning models require large-scale training to perform confidently, but obtaining annotated datasets in medical imaging is challenging. Weak annotation has emerged as a way to save time and effort. PURPOSE: To develop a deep learning model for 3D breast cancer segmentation in dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) using weak annotation with reliable performance. STUDY TYPE: Retrospective. POPULATION: Seven hundred and thirty-six women with breast cancer from a single institution, divided into the development (N = 544) and test dataset (N = 192). FIELD STRENGTH/SEQUENCE: 3.0-T, 3D fat-saturated gradient-echo axial T1-weighted flash 3D volumetric interpolated brain examination (VIBE) sequences. ASSESSMENT: Two radiologists performed a weak annotation of the ground truth using bounding boxes. Based on this, the ground truth annotation was completed through autonomic and manual correction. The deep learning model using 3D U-Net transformer (UNETR) was trained with this annotated dataset. The segmentation results of the test set were analyzed by quantitative and qualitative methods, and the regions were divided into whole breast and region of interest (ROI) within the bounding box. STATISTICAL TESTS: As a quantitative method, we used the Dice similarity coefficient to evaluate the segmentation result. The volume correlation with the ground truth was evaluated with the Spearman correlation coefficient. Qualitatively, three readers independently evaluated the visual score in four scales. A P-value <0.05 was considered statistically significant. RESULTS: The deep learning model we developed achieved a median Dice similarity score of 0.75 and 0.89 for the whole breast and ROI, respectively. The volume correlation coefficient with respect to the ground truth volume was 0.82 and 0.86 for the whole breast and ROI, respectively. The mean visual score, as evaluated by three readers, was 3.4. DATA CONCLUSION: The proposed deep learning model with weak annotation may show good performance for 3D segmentations of breast cancer using DCE-MRI. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY: Stage 2.

Fully Automatic Assessment of Background Parenchymal Enhancement on Breast MRI Using Machine-Learning Models.

BACKGROUND: Automated measurement and classification models with objectivity and reproducibility are required for accurate evaluation of the breast cancer risk of fibroglandular tissue (FGT) and background parenchymal enhancement (BPE). PURPOSE: To develop and evaluate a machine-learning algorithm for breast FGT segmentation and BPE classification. STUDY TYPE: Retrospective. POPULATION: A total of 794 patients with breast cancer, 594 patients assigned to the development set, and 200 patients to the test set. FIELD STRENGTH/SEQUENCE: 3T and 1.5T; T2 -weighted, fat-saturated T1 -weighted (T1 W) with dynamic contrast enhancement (DCE). ASSESSMENT: Manual segmentation was performed for the whole breast and FGT regions in the contralateral breast. The BPE region was determined by thresholding using the subtraction of the pre- and postcontrast T1 W images and the segmented FGT mask. Two radiologists independently assessed the categories of FGT and BPE. A deep-learning-based algorithm was designed to segment and measure the volume of whole breast and FGT and classify the grade of BPE. STATISTICAL TESTS: Dice similarity coefficients (DSC) and Spearman correlation analysis were used to compare the volumes from the manual and deep-learning-based segmentations. Kappa statistics were used for agreement analysis. Comparison of area under the receiver operating characteristic (ROC) curves (AUC) and F1 scores were calculated to evaluate the performance of BPE classification. RESULTS: The mean (±SD) DSC for manual and deep-learning segmentations was 0.85 ± 0.11. The correlation coefficient for FGT volume from manual- and deep-learning-based segmentations was 0.93. Overall accuracy of manual segmentation and deep-learning segmentation in BPE classification task was 66% and 67%, respectively. For binary categorization of BPE grade (minimal/mild vs. moderate/marked), overall accuracy increased to 91.5% in manual segmentation and 90.5% in deep-learning segmentation; the AUC was 0.93 in both methods. DATA CONCLUSION: This deep-learning-based algorithm can provide reliable segmentation and classification results for BPE. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY STAGE: 2.

Impact of antiviral treatment on long-term prognosis in non-immunocompromised patients with CMV reactivation.

BACKGROUND: Reactivation of human cytomegalovirus (CMV) occurs in non-immunocompromised patients with or without specific organ involvement, but it is still unknown whether it has a clinical implication on long-term prognosis or not. METHODS: A retrospective cohort study evaluating non-immunocompromised adult patients with CMV reactivation was conducted during the period between January 2010 and February 2018. Patients were divided into ganciclovir-treated and non-treated groups. Patients who died within 30 days from CMV reactivation were excluded as they died from complex causes of conditions. Survivors were followed for 30-months to evaluate long-term prognosis. RESULTS: A total of 136 patients with CMV reactivation was included, consisting of 66 ganciclovir-treated (48.5%) and 70 non-treated (51.5%) patients. Overall, patients were old-aged (median 70 years old) and most were treated with pneumonia of any cause (91.2%). More patients in ganciclovir-treated group were treated at intensive care unit (43.9% vs 24.3%, respectively) and had higher viral load over 5000 copies/ml (48.5% vs 22.9%) than non-treated group (all P < 0.05). Primary and secondary endpoints including 30-months survival (28.0 vs 38.9%, respectively) and 12-months survival (40.3% vs 49.2%) were not statistically different between the ganciclovir-treated and non-treated groups. In the multivariate analyses, ganciclovir treatment was not associated with 30-months survival (HR 1.307, 95% CI 0.759-2.251) and 12-months survival (HR 1.533, 95% CI 0.895-2.624). CONCLUSION: In a retrospective cohort study evaluating non-immunocompromised patients with CMV reactivation, ganciclovir treatment was not associated with long-term prognosis. Antiviral treatment in this condition would not be necessary unless organ involvement is suspected.

Risk Stratification of Ductal Carcinoma In Situ and Texture Analysis of Contrast-Enhanced Breast Magnetic Resonance Imaging.

OBJECTIVE: The aim of this study was to investigate whether texture analysis of contrast-enhanced T1 weighted images could predict risk of ductal carcinoma in situ (DCIS). METHODS: The study included 185 DCIS lesions that were classified as either low risk or non-low risk using surgical pathology records. All magnetic resonance imaging texture analyses were performed using postprocessing software, and texture-derived parameters were extracted. RESULTS: The sphericity, compactness, and spherical disproportion were significantly different in the low-risk and non-low risk groups using the Van Nuys Prognostic Index (mean ± SD, 0.479 ± 0.189 vs 0.414 ± 0.176, 0.161 ± 0.159 vs 0.112 ± 0.134, and 2.569 ± 1.434 vs 2.934 ± 1.374, respectively; P < 0.05). In the univariate analyses, sphericity (odds ratio, 7.091; 95% confidence interval, 1.236-40.666; P = 0.028) and compactness (odds ratio, 9.267; 95% confidence interval, 1.125-76.360; P = 0.039) were significantly associated with a high probability of being low risk according to the Van Nuys Prognostic Index. CONCLUSIONS: Whole-lesion texture analysis may be helpful in identifying patients classified as having low-risk DCIS before surgery.

Clinical factors influencing the performance of bacterial multiplex polymerase chain reaction in patients with community-onset pneumonia.

The etiologic diagnostic yield of community-onset pneumonia (COP) using conventional methods is low. Bacterial multiplex polymerase chain reaction (mPCR) has been shown to be more sensitive than conventional methods. This study assessed the clinical factors influencing bacterial mPCR results in patients with COP. Patients with COP admitted to a tertiary care hospital between November 2015 and April 2016 were retrospectively assessed. Conventional methods included culture-based methods and serology for Mycoplasma pneumoniae. Bacterial mPCR that could identify Streptococcus pneumoniae, Haemophilus influenzae, Mycoplasma pneumoniae, and Legionella pneumophilia was performed. Bacterial mPCR was performed in a total of 342 patients with COP in the study. Bacterial mPCR alone provided etiology in 99 patients. The total etiologic diagnosis rates improved from 22.2 to 51.1% when bacterial mPCR was added to conventional methods. Additional diagnostic benefits of bacterial mPCR were more prominent in the prior antibiotic non-exposure group (77.8% vs 63.5%, P = 0.015) and in the low-risk group with low CURB 65 score (62.6% vs 44.9%, P = 0.005). Patients who required ICU care, those with healthcare-associated pneumonia (HCAP), and patients with any underlying diseases were not associated with the additional pathogen detection rates using bacterial mPCR. By supplementing conventional diagnostic methods with bacterial mPCR-based methods, the overall pathogen detection rates improved in patients with COP. Moreover, the additional diagnostic usefulness of bacterial mPCR was significantly higher in patients without prior antibiotic exposure and in the mild-to-moderate-risk group with lower CURB 65 score.

Comparison of Positive Predictive Values of Categorization of Suspicious Calcifications Using the 4th and 5th Editions of BI-RADS.

OBJECTIVE. The objective of this study was to evaluate changes in the positive predictive value (PPV) of categorization of suspicious calcification for malignancy with the 4th versus the 5th edition of BI-RADS. MATERIALS AND METHODS. A total of 469 cases from 444 women (mean age, 50.1 years; age range, 23-82 years) with pathologically confirmed suspicious calcifications from January 2012 to June 2016 were enrolled in this retrospective study. Two radiologists determined morphology and distribution by consensus and categorized suspicious calcifications using the categorization systems in the 4th and 5th editions of BI-RADS. The PPVs for morphology, distribution, and categorization of calcifications were analyzed. The Pearson chi-square test was used to compare PPVs for morphology and distribution of suspicious calcification. RESULTS. The PPVs of categorization using the 5th edition matched better with BI-RADS category assessment than did categorization using the 4th edition. The PPVs of morphology were as follows: amorphous, 15.9%; coarse heterogeneous, 31.7%; fine pleomorphic, 58.2%; and fine linear or branching, 90.6% (p < 0.001). The PPVs of distribution were as follows: regional, 31.5%; grouped, 31.3%; linear, 50%; and segmental, 77.9% (p < 0.001). When distribution was divided into two types (grouped or regional vs segmental or linear) and analyzed in combination with suspicious morphology, distribution affected the PPVs and categorization of coarse heterogeneous, fine pleomorphic, and fine linear or branching calcifications. CONCLUSION. Categorization using both morphology and distribution according to the BI-RADS 5th edition was helpful to stratify risk levels of areas of suspicious calcification.

Host susceptibility to MERS-CoV infection, a retrospective cohort study of the 2015 Korean MERS outbreak.

To evaluate host susceptibility factors to Middle East respiratory syndrome coronavirus (MERS-CoV) infection, we conducted a retrospective cohort study from the single largest exposure event of the 2015 Korean MERS outbreak. A total of 175 patients were closely exposed to a super-spreader, 26 of which were infected (14.9%). In a multivariate analysis, history of autologous stem cell transplantation (HR, 31.151; 95% CI, 5.447-178.145; P < 0.001) and tachypnea at ED (HR, 4.392; 95% CI, 1.402-13.761; P = 0.011) were significantly associated with MERS-CoV infection.

MERS-CoV outbreak following a single patient exposure in an emergency room in South Korea: an epidemiological outbreak study.

BACKGROUND: In 2015, a large outbreak of Middle East respiratory syndrome coronavirus (MERS-CoV) infection occurred following a single patient exposure in an emergency room at the Samsung Medical Center, a tertiary-care hospital in Seoul, South Korea. We aimed to investigate the epidemiology of MERS-CoV outbreak in our hospital. METHODS: We identified all patients and health-care workers who had been in the emergency room with the index case between May 27 and May 29, 2015. Patients were categorised on the basis of their exposure in the emergency room: in the same zone as the index case (group A), in different zones except for overlap at the registration area or the radiology suite (group B), and in different zones (group C). We documented cases of MERS-CoV infection, confirmed by real-time PCR testing of sputum samples. We analysed attack rates, incubation periods of the virus, and risk factors for transmission. FINDINGS: 675 patients and 218 health-care workers were identified as contacts. MERS-CoV infection was confirmed in 82 individuals (33 patients, eight health-care workers, and 41 visitors). The attack rate was highest in group A (20% [23/117] vs 5% [3/58] in group B vs 1% [4/500] in group C; p<0·0001), and was 2% (5/218) in health-care workers. After excluding nine cases (because of inability to determine the date of symptom onset in six cases and lack of data from three visitors), the median incubation period was 7 days (range 2-17, IQR 5-10). The median incubation period was significantly shorter in group A than in group C (5 days [IQR 4-8] vs 11 days [6-12]; p<0·0001). There were no confirmed cases in patients and visitors who visited the emergency room on May 29 and who were exposed only to potentially contaminated environment without direct contact with the index case. The main risk factor for transmission of MERS-CoV was the location of exposure. INTERPRETATION: Our results showed increased transmission potential of MERS-CoV from a single patient in an overcrowded emergency room and provide compelling evidence that health-care facilities worldwide need to be prepared for emerging infectious diseases. FUNDING: None.

Atypical presentations of MERS-CoV infection in immunocompromised hosts.

During the 2015 Korean MERS outbreak, we experienced atypical presentations of MERS-CoV infections in three immunocompromised hosts that warranted exceptional management. Case 1 showed delayed symptom development after a four-day asymptomatic period, Case 2 experienced a 20-day incubation period, and Case 3 exhibited persistent viral shedding without clinical deterioration. Recognizing these exceptions is extremely important in the management of MERS-CoV-exposed or -infected patients and for control of potential MERS outbreaks.

Histogram analysis of volume-based apparent diffusion coefficient in breast cancer.

Background Breast cancer is a heterogeneous disease. Recent studies showed that apparent diffusion coefficient (ADC) values have various association with tumor aggressiveness and prognosis. Purpose To evaluate the value of histogram analysis of ADC values obtained from the whole tumor volume in invasive ductal cancer (IDC) and ductal carcinoma in situ (DCIS). Material and Methods This retrospective study included 201 patients with confirmed DCIS (n = 37) and IDC (n = 164). The IDC group was divided into two groups based on the presence of a DCIS component: IDC-DCIS (n = 76) and pure IDC (n = 88). All patients underwent preoperative breast magnetic resonance imaging (MRI) with diffusion-weighted images at 3.0 T. Histogram parameters of cumulative ADC values, skewness, and kurtosis were calculated and statistically analyzed. Results The differences between DCIS, IDC-DCIS, and pure IDC were significant in all percentiles of ADC values, in descending order of DCIS, IDC-DCIS, and pure IDC. IDC showed significantly lower ADC values than DCIS, and ADC50 was the best indicator for discriminating IDC from DCIS, with a threshold of 1.185 × 10-3 mm2/s (sensitivity of 82.9%, specificity of 75.7%). However, multivariate analysis of obtained ADC values showed no significant differences between DCIS, IDC-DCIS, and pure IDC ( P > 0.05). Conclusion Volume-based ADC values showed association with heterogeneity of breast cancer. However, there was no additional diagnostic performance in histogram analysis for differentiating between DCIS, IDC-DCIS, and pure IDC.

Control of an Outbreak of Middle East Respiratory Syndrome in a Tertiary Hospital in Korea.

BACKGROUND: In 2015, a large outbreak of Middle East respiratory syndrome (MERS) occurred in the Republic of Korea. Half of the cases were associated with a tertiary care university hospital. OBJECTIVE: To document the outbreak and successful control measures. DESIGN: Descriptive study. SETTING: A 1950-bed tertiary care university hospital. PATIENTS: 92 patients with laboratory-confirmed MERS and 9793 exposed persons. MEASUREMENTS: Description of the outbreak, including a timeline, and evaluation of the effectiveness of the control measures. RESULTS: During the outbreak, 92 laboratory-confirmed MERS cases were associated with a large tertiary care hospital, 82 of which originated from unprotected exposure to 1 secondary patient. Contact tracing and monitoring exposed patients and assigned health care workers were at the core of the control measures in the outbreak. Nontargeted screening measures, including body temperature screening among employees and visitors at hospital gates, monitoring patients for MERS-related symptoms, chest radiographic screening, and employee symptom monitoring, did not detect additional patients with MERS without existing transmission links. All in-hospital transmissions originated from 3 patients with MERS who also had pneumonia and productive cough. LIMITATIONS: This was a retrospective single-center study. Statistical analysis could not be done. Because this MERS outbreak originated from a superspreader, effective control measures could differ in endemic areas or in other settings. CONCLUSION: Control strategies for MERS outbreaks should focus on tracing contacts of persons with epidemiologic links. Adjusting levels of quarantine and personal protective equipment according to the assumed infectivity of each patient with MERS may be appropriate. PRIMARY FUNDING SOURCE: Samsung Biomedical Research Institute.

Differential Cell Count and CRP Level in Blood as Predictors for Middle East Respiratory Syndrome Coronavirus Infection in Acute Febrile Patients during Nosocomial Outbreak.

A case-control study was performed to identify clinical predictors for Middle East respiratory syndrome coronavirus (MERS-CoV) infection among patients with acute febrile illness during the nosocomial outbreak. Patients with MERS-CoV were more likely to have monocytosis with normal white blood cell (WBC) count and lower C-reactive protein (CRP) level. Simple laboratory data such as complete blood counts (CBC) with differential count could be a useful marker for the prediction of MERS and triage at the initial presentation of acute febrile patients in outbreak setting.

Clinical Features and Risk Factors for Development of Breakthrough Gram-Negative Bacteremia during Carbapenem Therapy.

With the increasing use of carbapenems, carbapenem-resistant Gram-negative bacteria have become a major concern in health care-associated infections. The present study was performed to evaluate the clinical and microbiological features of breakthrough Gram-negative bacteremia (GNB) during carbapenem therapy and to assess risk factors for development of breakthrough GNB. A case-control study was performed at a tertiary hospital from 2005 to 2014. Case patients were defined as individuals whose blood cultures grew Gram-negative bacteria while the patients were receiving carbapenems for at least 48 h before breakthrough GNB. Age-, sex-, and date-matched controls were selected from patients who received carbapenem for at least 48 h and did not develop breakthrough GNB during carbapenem treatment. A total of 101 cases of breakthrough GNB were identified and compared to 100 controls. The causative microorganisms for breakthrough GNB were Stenotrophomonas maltophilia (n = 33), Acinetobacter baumannii (n = 32), Pseudomonas aeruginosa (n = 21), and others (n = 15). Approximately 90% of S. maltophilia isolates were susceptible to levofloxacin and trimethoprim-sulfamethoxazole. The most common infection types were primary bacteremia (38.6%) and respiratory infections (35.6%). More than half of the patients died within a week after bacteremia, and the 30-day mortality rate was 70.3%. In a multivariate analysis, a longer hospital stay, hematologic malignancy, persistent neutropenia, immunosuppressant use, and previous colonization by causative microorganisms were significantly associated with breakthrough GNB. Our data suggest that S. maltophilia, A. baumannii, and P. aeruginosa are the major pathogens of breakthrough GNB during carbapenem therapy, in association with a longer hospital stay, hematologic malignancy, persistent neutropenia, immunosuppressant use, and previous colonization.

Case-control study of the risk factors for acquisition of Pseudomonas and Proteus species during tigecycline therapy.

Tigecycline is an important agent in clinical practice because of its broad-spectrum activity. However, it has no activity against Pseudomonas or Proteus species. We conducted a case-control study to analyze risk factors for the acquisition of Pseudomonas or Proteus spp. during tigecycline therapy. Placement of suction drainage at infected wound sites, ICU stay, and neurologic disease were identified as independent risk factors for the acquisition of Pseudomonas and Proteus spp.

Histogram analysis of apparent diffusion coefficient at 3.0t: Correlation with prognostic factors and subtypes of invasive ductal carcinoma.

PURPOSE: To evaluate apparent diffusion coefficient (ADC) histogram parameters that show correlations with prognostic factors and subtypes of breast cancer. MATERIALS AND METHODS: At 3.0T, various ADC histogram parameters were calculated including the entire tumor volume in 173 invasive ductal carcinomas: the minimum, 10th percentile, mean, median, 90th percentile, and maximum. ADC parameters were correlated with prognostic factors and subtype. RESULTS: The mean ADCmedian value was significantly higher in the group with lymph node metastasis, HER2 positivity, and a Ki-67 value <14% than in the group with negativity for lymph node metastasis, HER2 negativity, and a Ki-67 value ≥14% (0.907, 0.978, and 0.941 vs. 0.735, 0.778, and 0.761 × 10(-3) mm(2) /s, respectively) (P < 0.01). There was no significant correlation between ADCmedian and tumor size, histologic grade, estrogen receptor expression, and progesterone receptor expression (P = 0.272, 0.113, 0.261, and 0.181, respectively). For most ADC parameters except for ADCmin , the mean of variable ADC parameters of HER2-positive, luminal A, luminal B-HER2(+), triple-negative, and luminal B-HER2(-) diseases were arranged in descending order (1.175, 0.936, 0.863, 0.811, and 0.665 × 10(-3) mm(2) /s in ADCmedian , respectively) with statistical significant difference (P < 0.001). In multivariate analysis, histologic grade, the Ki-67 index, and HER2 expression were statistically significant explanatory prognostic factors for ADCmedian and the Ki-67 index had the most robust effects on ADC parameters (standardized coefficient = -0.317). CONCLUSION: Various ADC parameters were correlated with prognostic factors and subtype, except for ADCmin . HER2 positivity showed high ADC values and high Ki-67 index revealed low ADC values.

A Large Cohort Study of Height and Mammographic Density in Relation to Breast Cancer Risk among Korean Women.

BACKGROUND: Height and mammographic breast density are well-known risk factors for breast cancer. This study aims to investigate the association between height and mammographic density with breast cancer risk in a large population-based cohort of Korean women. METHODS: This retrospective cohort study included 4,851,115 women aged 40 and older who underwent screening mammography through the Korean National Cancer Screening Program (KNCSP) between 2009 and 2014 and were followed up until 2016. Multivariable Cox proportional hazard models were employed to estimate adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for the associations between height, mammographic density, and breast cancer risk. RESULTS: A taller stature was associated with an increased risk of breast cancer, with women in the highest quintile of height (Q5) having a 1.54-fold [95% CI: 1.49, 1.59] greater risk compared to those in the lowest quintile (Q1). When analyzing breast cancer risk based on height and mammographic density, women in the highest quintile for height (Q5) with extremely dense breasts had a 4.51-fold [95% CI: 4.24, 4.79] greater risk compared to those in the lowest quintile (Q1) with almost entirely fatty breasts (p trend < 0.001). CONCLUSION: This first study to simultaneously examine these two variables found that height and mammographic density were independently associated with breast cancer risk in this large cohort of Korean women. IMPACT: Height and mammographic density can help stratify risk in screening populations for breast cancer. Careful consideration of screening strategies may be beneficial for taller women with dense breasts.

The Role of Diffusion-Weighted Imaging Based on Maximum-Intensity Projection in Young Patients with Marked Background Parenchymal Enhancement on Contrast-Enhanced Breast MRI.

Diffusion-weighted imaging (DWI) utilizing maximum-intensity projection (MIP) was suggested as a cost-effective alternative tool without the risk of gadolinium-based contrast agents. The purpose of this study was to investigate whether DWI MIPs played a supportive role in young (≤60) patients with marked background parenchymal enhancement (BPE) on contrast-enhanced MRI (CE-MRI). The research included 1303 patients with varying degrees of BPE, and correlations between BPE on CE-MRI, the background diffusion signal (BDS) on DWI, and clinical parameters were analyzed. Lesion detection scores were compared between CE-MRI and DWI, with DWI showing higher scores. Among the 186 lesions in 181 patients with marked BPE on CE-MRI, the main lesion on MIPs of CE-MRI was partially or completely seen in 88.7% of cases, while it was not seen in 11.3% of cases. On the other hand, the main lesion on MIPs of DWI was seen in 91.4% of cases, with only 8.6% of cases showing no visibility. DWI achieved higher scores for lesion detection compared to CE-MRI. The presence of a marked BDS was significantly associated with a lower likelihood of a higher DWI score (p < 0.001), and non-mass lesions were associated with a decreased likelihood of a higher DWI score compared with mass lesions (p = 0.196). In conclusion, the inclusion of MIPs of DWI in the preoperative evaluation of breast cancer patients, particularly young women with marked BPE, proved highly beneficial in improving the overall diagnostic process.

[MRI-Guided Breast Intervention: Biopsy and Needle Localization]. / ìžê¸°ê³µëª ì˜ìƒ ìœ ë„í•˜ ìœ ë°©ì˜ ì¤‘ìž¬ì ì‹œìˆ : ì¡°ì§ìƒê²€ìˆ ë° ì¹¨ìœ„ì¹˜ê²°ì •ìˆ .

In Korea, the number of institutions providing breast MRI, as well as the number of breast MRIs, has recently increased. However, MRI-guided procedures, including biopsy and needle localization, are rarely performed compared to ultrasound-guided or stereotactic biopsy. As breast MRI has high sensitivity but limited specificity, lesions detected only on MRI require pathologic confirmation through MRI-guided biopsy or surgical excision with MRI-guided needle localization. Thus, we aimed to review MRI-guided procedures, including their indications, techniques, procedural considerations, and limitations.

Diagnostic Usefulness of Diffusion-Weighted MRI for Axillary Lymph Node Evaluation in Patients with Breast Cancer.

This study aimed to determine whether apparent diffusion coefficient (ADC) and morphological features on diffusion-weighted MRI (DW-MRI) can discriminate metastatic axillary lymph nodes (ALNs) from benign in patients with breast cancer. Two radiologists measured ADC, long and short diameters, long-to-short diameter ratio, and cortical thickness and assessed eccentric cortical thickening, loss of fatty hilum, irregular margin, asymmetry in shape or number, and rim sign of ALNs on DW-MRI and categorized them into benign or suspicious ALNs. Pathologic reports were used as a reference standard. Statistical analysis was performed using the Mann-Whitney U test and chi-square test. Overall sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and diagnostic accuracy of DW-MRI were calculated. The ADC of metastatic ALNs was 0.905 × 10-3 mm2/s, and that of benign ALNs was 0.991 × 10-3 mm2/s (p = 0.243). All morphologic features showed significant difference between the two groups. The sensitivity, specificity, PPV, NPV, and diagnostic accuracy of the final categorization on DW-MRI were 77.1%, 93.3%, 79.4%, 92.5%, and 86.2%, respectively. Our results suggest that morphologic evaluation of ALNs on DWI can discriminate metastatic ALNs from benign. The ADC value of metastatic ALNs was lower than that of benign nodes, but the difference was not statistically significant.