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.

Emerging trends of inflammatory bowel disease in South Korea: A nationwide population-based study.

BACKGROUND AND AIM: Little is known regarding the exact burden of inflammatory bowel disease (IBD) in Asian countries because previous epidemiologic studies were hospital based. We aimed to develop and validate an operational definition of IBD cases from health insurance claims data and to examine the epidemiological features of IBD in Korea. METHODS: We analyzed stratified sample data from the Korean Health Insurance Review and Assessment (2010-2016) database using 12 different definitions and applied the best definition to the entire (2007-2016) dataset. RESULTS: The definition that combined the International Classification of Disease 10th revision code with IBD-specific medications had the best performance characteristics among the 12 tested definitions. During the 8-year study period, IBD prevalence increased from 25 345 in 2009 to 47 444 in 2016. Over that period, the prevalence of Crohn's disease increased 1.9-fold (from 16.0/100 000 in 2009 to 29.6/100 000 in 2016) and that of ulcerative colitis increased 1.6-fold (from 41.4/100 000 in 2009 to 66.0/100 000 in 2016). Similarly, the estimated incidence of Crohn's disease also increased 1.2-fold (from 2.4 to 2.9 per 100 000) and that of ulcerative colitis rose 1.3-fold (from 4.0 to 5.3 per 100 000). During the study period, the predominant increase in IBD incidence was among younger individuals, especially those aged < 30 years. CONCLUSION: Patients with IBD can be accurately identified using Korean insurance claims data by combining information regarding the International Classification of Disease 10th revision codes and the IBD medications used. The prevalence of IBD continues to increase, with an apparent shift toward younger (< 30 years) age groups.

Assessment of the root apex position of impacted maxillary canines on panoramic films.

INTRODUCTION: In this study, we aimed to analyze the root apex positions of impacted maxillary canines on panoramic radiographs according to their labiopalatal positions and to propose a new panoramic radiographic predictor of impacted maxillary canines. METHODS: Paired panoramic films and cone-beam computed tomography images of 104 patients with unilateral impacted maxillary canines were selected. The sector locations of the impacted canine root apices on the panoramic radiographs were compared with the labiopalatal positions of impacted maxillary canines on cone-beam computed tomography. Statistical correlations between panoramic and cone-beam computed tomography findings were examined with the chi-square test. RESULTS: A statistically significant association was observed between the sector locations of impacted canine root apices and the labiopalatal positions of the canines (P <0.001). The root apices of labially impacted canines were more frequently located in sector 2 (corresponding to the lateral incisor position), those of palatally impacted canines were more frequently located in sector 4 (corresponding to the first premolar), and those of midalveolus impacted canines were more frequently located in sector 3 (corresponding to the canine). CONCLUSIONS: On panoramic films, the root apices of palatally impacted canines tended toward the root apices of the maxillary first premolars, whereas those of labially impacted canines tended toward the root apices of the maxillary lateral incisors. Therefore, the sector locations of impacted canine root apices on panoramic radiographs could be used for the early prediction of maxillary canine impactions and determination of the labiopalatal positions of impacted canines.

Clinical usefulness of the free web-based image analysis application ImmunoRatio for assessment of Ki-67 labelling index in breast cancer.

AIMS: Ki-67 is a prognostic marker in breast cancer; however, the use of the Ki-67 labelling index (LI) in clinical practice requires a consistent and easily accessible scoring method. The present study evaluated the use of the free internet-based image analysis program ImmunoRatio to score Ki-67 LI in breast cancer in comparison with manual counting. METHODS: Ki-67 immunohistochemical detection was performed in 577 breast cancer cases, and the Ki-67 LI was determined by ImmunoRatio and manual counting. RESULTS: The Ki-67 LI determined by ImmunoRatio correlated well with that obtained by manual counting. The concordance rate between ImmunoRatio and manual counting was excellent (κ coefficient of 0.881) at a Ki-67 LI cut-off value of 20%. Cases with high Ki-67 LI by ImmunoRatio were associated with poor overall survival, in particular in the hormone receptor positive group. CONCLUSIONS: The web-based automated image analysis program ImmunoRatio is an attractive alternative to manual counting to determine the Ki-67 LI in breast cancer.

Breast Magnetic Resonance Imaging for Assessment of Internal Mammary Lymph Node Status in Breast Cancer.

PURPOSE: The purpose of this study was to assess magnetic resonance imaging (MRI) features of malignant internal mammary lymph nodes (IMLNs) and benign IMLNs in breast cancer patients. METHODS: From 2009 to 2014, the records of 85 patients with IMLNs were archived using MRI report data; 26 patients with small size (long axis diameter <5 mm) nodes were subsequently excluded. The current study evaluated internal mammary lymph nodes in 59 patients who underwent breast MRI for breast cancer staging and for posttherapy follow-up. All MRI findings were retrospectively evaluated. Malignancy was determined based on pathologic examination and positron emission tomography computed tomography findings. Independent t-tests, Mann-Whitney U tests, chi-square tests, and receiver operating characteristics (ROC) curve analysis were used. RESULTS: Among MRI features, there were statistically significant differences between benign and malignant IMLN groups, in short axis length (3.6±1.3 vs. 8.2±2.9 mm, respectively), long axis length (8.1±2.4 vs. 14.5±4.8 mm, respectively), short/long axis ratio (0.45±0.10 vs. 0.59±0.17, respectively), absent fatty hilum (mean, 0% vs. 95%, respectively), and restricted diffusion (15.8% vs. 85.0%, respectively) (p<0.050). Multiplicity and location of intercostal spaces was not different between the two groups. Short axis length was the most discriminative variable for predicting metastatic nodes (area under the ROC curve, 0.951; threshold, 4 mm; sensitivity, 92.5%; specificity, 84.2%). CONCLUSION: Conventional MRI and diffusion-weighted MRI are helpful to detect metastasis of internal mammary lymph nodes in breast cancer.

Monitoring by LC-MS/MS of 48 compounds of sildenafil, tadalafil, vardenafil and their analogues in illicit health food products in the Korean market advertised as enhancing male sexual performance.

More than 46 phosphodiesterase type 5 (PDE5) inhibitor analogues have been found to be present as illegal adulterants in various forms of health food products (powder, tablet, capsule, etc.), thereby placing the health of consumers at risk through product intake. In this study, 164 samples advertised to be effective at enhancing male sexual performance were collected over a 4-year period (2009-2012) from the Korean on-line or off-line market and screened. An LC-MS/MS method was employed to screen for the presence of 48 compounds including sildenafil, tadalafil, vardenafil and their analogues. Method validation established LOQs (0.30-10.00 ng ml(-1) or ng g(-1)) and recoveries (spiked in liquid sample, 84-112%; spiked in solid sample, 83-110%). Most of the illicit products screened were adulterated with 14 of the PDE5 derivatives under examination, including considerable amounts of sildenafil and tadalafil; of the 48 compounds, tadalafil was the most frequent adulterant (42.6%), followed by sildenafil (27.9%). Specifically, tadalafil concentration ranges (mg g(-1)) in the samples collected over the 4-year period were determined as follows: 2.91-52.20 (2009), 4.50-108.10 (2010), 0.37-101.40 (2011), and 0.08-138.69 mg g(-1) (2012). The concentration ranges (mg g(-1)) of sildenafil were also at high levels: 4.90-117.96 (2009), 1.30-369.93 (2010), 0.03-241.77 (2011), and 18.34-297.91 mg g(-1) (2012). The results of screening for PDE5 inhibitor pharmaceuticals as adulterants in illicit health food products are of great significance with respect to the protection of public health and consumer safety.

Histological assessment of mechanoreceptors in Achilles allografts after anterior cruciate ligament reconstruction.

BACKGROUND: There is a lack of histological studies investigating the presence of mechanoreceptors in anterior cruciate ligament (ACL) allografts. HYPOTHESIS: Mechanoreceptors would not grow in Achilles allografts after ACL reconstruction. STUDY DESIGN: Case series study; Level of evidence, 4. METHODS: Tissue samples were obtained from 11 patients who underwent ACL reconstruction using Achilles tendon allografts. They underwent biopsies during second-look arthroscopies. The mean period from ACL reconstruction to harvesting tissue was 26.63 months (range, 12-120 months). The control group consisted of 2 normal ACLs procured from 42- and 45-year-old men who underwent amputation above the knee due to trauma. RESULTS: Ruffini corpuscles and free nerve endings were shown to be present in the specimens of the control group by processing hematoxylin-eosin stains and immunohistochemical stains with monoclonal antibodies against S-100. In the Achilles allografts, mechanoreceptors were not observed. However, fibroblasts, collagen fibers, and vessels that were not present in fresh-frozen Achilles allografts before surgery were observed. CONCLUSION: The results demonstrate that Achilles tendon allografts appeared similar to normal ligaments except for the lack of histological evidence of mechanoreceptors. In other words, there are no newly ingrown mechanoreceptors in ACL allografts.