Explore the vast range of titles available.

PurposeTo develop a deep learning–based computer-aided diagnosis (CAD) system for use in the CT diagnosis of cervical lymph node metastasis (LNM) in patients with thyroid cancer.MethodsA total of 995 axial CT images that included benign (n = 647) and malignant (n = 348) lymph nodes were collected from 202 patients with thyroid cancer who underwent CT for surgical planning between July 2017 and January 2018. The datasets were randomly split into training (79.0%), validation (10.5%), and test (10.5%) datasets. Eight deep convolutional neural network (CNN) models were used to classify the images into metastatic or benign lymph nodes. Pretrained networks were used on the ImageNet and the best-performing algorithm was selected. Class-specific discriminative regions were visualized with attention heatmap using a global average pooling method.ResultsThe area under the ROC curve (AUROC) for the tested algorithms ranged from 0.909 to 0.953. The sensitivity, specificity, and accuracy of the best-performing algorithm were all 90.4%, respectively. Attention heatmap highlighted important subregions for further clinical review.ConclusionA deep learning–based CAD system could accurately classify cervical LNM in patients with thyroid cancer on preoperative CT with an AUROC of 0.953. Whether this approach has clinical utility will require evaluation in a clinical setting.Key Points• A deep learning–based CAD system could accurately classify cervical lymph node metastasis. The AUROC for the eight tested algorithms ranged from 0.909 to 0.953.• Of the eight models, the ResNet50 algorithm was the best-performing model for the validation dataset with 0.953 AUROC. The sensitivity, specificity, and accuracy of the ResNet50 model were all 90.4%, respectively, in the test dataset.• Based on its high accuracy of 90.4%, we consider that this model may be useful in a clinical setting to detect LNM on preoperative CT in patients with thyroid cancer.

The prevalence and clinical characteristics of depressive disorders differ between women and men; however, the genetic contribution to sex differences in depressive disorders has not been elucidated. To evaluate sex-specific differences in the genetic architecture of depression, whole exome sequencing of samples from 1000 patients (70.7% female) with depressive disorder was conducted. Control data from healthy individuals with no psychiatric disorder (n = 72, 26.4% female) and East-Asian subpopulation 1000 Genome Project data (n = 207, 50.7% female) were included. The genetic variation between men and women was directly compared using both qualitative and quantitative research designs. Qualitative analysis identified five genetic markers potentially associated with increased risk of depressive disorder in females, including three variants (rs201432982 within PDE4A , and rs62640397 and rs79442975 within FDX1L ) mapping to chromosome 19p13.2 and two novel variants (rs820182 and rs820148) within MYO15B at the chromosome 17p25.1 locus. Depressed patients homozygous for these variants showed more severe depressive symptoms and higher suicidality than those who were not homozygotes (i.e., heterozygotes and homozygotes for the non-associated allele). Quantitative analysis demonstrated that the genetic burden of protein-truncating and deleterious variants was higher in males than females, even after permutation testing. Our study provides novel genetic evidence that the higher prevalence of depressive disorders in women may be attributable to inherited variants.

Colposcopy is widely used to detect cervical cancers, but experienced physicians who are needed for an accurate diagnosis are lacking in developing countries. Artificial intelligence (AI) has been recently used in computer-aided diagnosis showing remarkable promise. In this study, we developed and validated deep learning models to automatically classify cervical neoplasms on colposcopic photographs. Pre-trained convolutional neural networks were fine-tuned for two grading systems: the cervical intraepithelial neoplasia (CIN) system and the lower anogenital squamous terminology (LAST) system. The multi-class classification accuracies of the networks for the CIN system in the test dataset were 48.6 ± 1.3% by Inception-Resnet-v2 and 51.7 ± 5.2% by Resnet-152. The accuracies for the LAST system were 71.8 ± 1.8% and 74.7 ± 1.8%, respectively. The area under the curve (AUC) for discriminating high-risk lesions from low-risk lesions by Resnet-152 was 0.781 ± 0.020 for the CIN system and 0.708 ± 0.024 for the LAST system. The lesions requiring biopsy were also detected efficiently (AUC, 0.947 ± 0.030 by Resnet-152), and presented meaningfully on attention maps. These results may indicate the potential of the application of AI for automated reading of colposcopic photographs.

The primary goal of this study was to evaluate the major roles of health-related quality of life (HRQOL) in a 5-year lung cancer survival prediction model using machine learning techniques (MLTs). The predictive performances of the models were compared with data from 809 survivors who underwent lung cancer surgery. Each of the modeling technique was applied to two feature sets: feature set 1 included clinical and sociodemographic variables, and feature set 2 added HRQOL factors to the variables from feature set 1. One of each developed prediction model was trained with the decision tree (DT), logistic regression (LR), bagging, random forest (RF), and adaptive boosting (AdaBoost) methods, and then, the best algorithm for modeling was determined. The models’ performances were compared using fivefold cross-validation. For feature set 1, there were no significant differences in model accuracies (ranging from 0.647 to 0.713). Among the models in feature set 2, the AdaBoost and RF models outperformed the other prognostic models [area under the curve (AUC) = 0.850, 0.898, 0.981, 0.966, and 0.949 for the DT, LR, bagging, RF and AdaBoost models, respectively] in the test set. Overall, 5-year disease-free lung cancer survival prediction models with MLTs that included HRQOL as well as clinical variables improved predictive performance.

In light of recent developments in genomic technology and the rapid accumulation of genomic information, a major transition toward precision medicine is anticipated. However, the clinical applications of genomic information remain limited. This lag can be attributed to several complex factors, including the knowledge gap between medical experts and bioinformaticians, the distance between bioinformatics workflows and clinical practice, and the unique characteristics of genomic data, which can make interpretation difficult. Here we present a novel genomic data model that allows for more interactive support in clinical decision-making. Informational modelling was used as a basis to design a communication scheme between sophisticated bioinformatics predictions and the representative data relevant to a clinical decision. This study was conducted by a multidisciplinary working group who carried out clinico-genomic workflow analysis and attribute extraction, through Failure Mode and Effects Analysis (FMEA). Based on those results, a clinical genome data model (cGDM) was developed with 8 entities and 46 attributes. The cGDM integrates reliability-related factors that enable clinicians to access the reliability problem of each individual genetic test result as clinical evidence. The proposed cGDM provides a data-layer infrastructure supporting the intellectual interplay between medical experts and informed decision-making.

Porcine respiratory coronavirus (PRCV) is a member of the species Alphacoronavirus 1 within the genus Alphacoronavirus of the family Coronaviridae. A few studies have been conducted on the prevalence of PRCV since its first identification in 1997, but there have been no recent studies on the prevalence and genetic characterization of the virus in Korea. In this study, the seroprevalence of PRCV was determined in Korean pig farms using a commercially available TGEV/PRCV differential enzyme-linked immunosorbent assay kit. The farm-level seroprevalence of PRCV was determined to be 68.6% (48/70), similar to previous reports in Korea, suggesting that PRCV is still circulating in Korean pig herds nationwide. Among the 20 PRCV-seropositive farms tested in this study, PRCV RNAs were detected in 17 oral fluid samples (28.3%) from nine farms (45.0%), while TGEV RNAs were not detected in any sample. To investigate the genetic characteristics of Korean PRCV strains, genetic and phylogenetic analyses were conducted on PRCV spike gene sequences obtained in this study. The three Korean PRCV strains (KPRCV2401, KPRCV2402, and KPRCV2403) shared 98.5–100% homology with each other and 96.2–96.6% and 91.6–94.5% homology with European and American strains, respectively. A 224-amino acid deletion was found in the S gene of both Korean and European PRCVs but not in that of American PRCVs, suggesting a European origin for Korean PRCVs. Phylogenetic analysis showed that Korean PRCVs are more closely related to European PRCVs than American PRCVs but clustered apart from both, suggesting that Korean PRCV has evolved independently since its emergence in Korean PRCVs. The results of this study will help expand knowledge on the epidemiology and molecular biology of PRCV currently circulating in Korea.

Home blood pressure measurement (HBPM) has the advantage of measuring blood pressure (BP) multiple times over a long period. HBPM effectively diagnoses stress-induced transient BP elevations (i.e., white coat hypertension), insufficient BP control throughout the day (i.e., masked hypertension), and even BP variability. In most cases, HBPM may increase self-awareness of BP, increasing the compliance of treatment. Cumulative evidence has reported better improved predictive values of HBPM in cardiovascular morbidity and mortality than office BP monitoring. In this position paper, the Korean Society of Hypertension Home Blood Pressure Forum provides comprehensive information and clinical importance on HBPM.

In recent years, there has been an emerging interest in the use of claims and electronic health record (EHR) data for evaluation of medical device safety and effectiveness. In Korea, national insurance electronic data interchange (EDI) code has been used as a medical device data source for common data model (CDM). This study performed a preliminary feasibility assessment of CDM-based vigilance. A cross-sectional study of target medical device data in EHR and CDM was conducted. A total of 155 medical devices were finally enrolled, with 58.7% of them having EDI codes. Femoral head prosthesis was selected as a focus group. It was registered in our institute with 11 EDI codes. However, only three EDI codes were converted to systematized nomenclature of medicine clinical terms concept. EDI code was matched in one-to-many (up to 104) with unique device identifier (UDI), including devices classified as different global medical device nomenclature. The use of UDI rather than EDI code as a medical device data source is recommended. We hope that this study will share the current state of medical device data recorded in the EHR and contribute to the introduction of CDM-based medical device vigilance by selecting appropriate medical device data sources.

Uveal melanoma(UM) is the most common primary intraocular malignancy in adults. However, the incidence of UM in Asia is 10 to 20 times less than in Western populations. Therefore, for the first time, we report our whole exome sequencing (WES) data analysis to discover differences in the molecular features of Asian and Western UM, and to determine the disparities between the primary tumor before brachytherapy and enucleated samples after brachytherapy. WES of 19 samples (13 primary tumors, 5 enucleation samples after brachytherapy, and 1 liver metastasis) from 13 patients diagnosed with UM and treated between 2007 and 2019 at the Yonsei University Health System (YUHS) were analyzed using bioinformatics pipelines. We identified significantly altered genes in Asian UM and changes in mutational profiles before and after brachytherapy using various algorithms. GNAQ, BAP1, GNA11, SF3B1 and CYSLTR2 were significantly mutated in Asian UM, which is similar that reported frequently in previous Western-based UM studies. There were also similar copy number alterations (M3, 1p loss, 6p gain, 8q gain) in both groups. In paired comparisons of the same patients, DICER1 and LRP1B were distinctly mutated only in tumor samples obtained after brachytherapy using rare-variant association tests ( P  = 0.01, 0.01, respectively). The mutational profiles of Asian UM were generally similar to the data from previous Western-based studies. DICER1 and LRP1B were newly mutated genes with statistical significance in the regrowth samples after brachytherapy compared to the primary tumors, which may be related to resistance to brachytherapy.

NUDT15 and TPMT variants are strong genetic determinants of thiopurine-induced hematological toxicity. Despite the impact of homozygous CRIM1 on thiopurine toxicity, several patients with wild-type NUDT15, TPMT, and CRIM1 experience thiopurine toxicity, therapeutic failure, and relapse of acute lymphoblastic leukemia (ALL). Novel pharmacogenetic interactions associated with thiopurine intolerance from hematological toxicities were investigated using whole-exome sequencing for last-cycle 6-mercaptopurine dose intensity percentages (DIP) tolerated by pediatric ALL patients ( N  = 320). IL6 rs13306435 carriers ( N  = 19) exhibited significantly lower DIP (48.0 ± 27.3%) than non-carriers ( N  = 209, 69.9 ± 29.0%; p  = 0.0016 and 0.0028 by t test and multiple linear regression, respectively). Among 19 carriers, 7 with both heterozygous IL6 rs13306435 and CRIM1 rs3821169 showed significantly decreased DIP (24.7 ± 8.9%) than those with IL6 ( N  = 12, 61.6 ± 25.1%) or CRIM1 ( N  = 94, 68.1 ± 28.4%) variants. IL6 and CRIM1 variants showed marked inter-ethnic variability. Four-gene-interplay models revealed the best odds ratio (8.06) and potential population impact [relative risk (5.73), population attributable fraction (58%), number needed to treat (3.67), and number needed to genotype (12.50)]. Interplay between IL6 rs13306435 and CRIM1 rs3821169 was suggested as an independent and/or additive genetic determinant of thiopurine intolerance beyond NUDT15 and TPMT in pediatric ALL .