Development and operation of a digital platform for sharing pathology image data.

BACKGROUND: Artificial intelligence (AI) research is highly dependent on the nature of the data available. With the steady increase of AI applications in the medical field, the demand for quality medical data is increasing significantly. We here describe the development of a platform for providing and sharing digital pathology data to AI researchers, and highlight challenges to overcome in operating a sustainable platform in conjunction with pathologists. METHODS: Over 3000 pathological slides from five organs (liver, colon, prostate, pancreas and biliary tract, and kidney) in histologically confirmed tumor cases by pathology departments at three hospitals were selected for the dataset. After digitalizing the slides, tumor areas were annotated and overlaid onto the images by pathologists as the ground truth for AI training. To reduce the pathologists' workload, AI-assisted annotation was established in collaboration with university AI teams. RESULTS: A web-based data sharing platform was developed to share massive pathological image data in 2019. This platform includes 3100 images, and 5 pre-processing algorithms for AI researchers to easily load images into their learning models. DISCUSSION: Due to different regulations among countries for privacy protection, when releasing internationally shared learning platforms, it is considered to be most prudent to obtain consent from patients during data acquisition. CONCLUSIONS: Despite limitations encountered during platform development and model training, the present medical image sharing platform can steadily fulfill the high demand of AI developers for quality data. This study is expected to help other researchers intending to generate similar platforms that are more effective and accessible in the future.

PAIP 2020: Microsatellite instability prediction in colorectal cancer.

Microsatellite instability (MSI) refers to alterations in the length of simple repetitive genomic sequences. MSI status serves as a prognostic and predictive factor in colorectal cancer. The MSI-high status is a good prognostic factor in stage II/III cancer, and predicts a lack of benefit to adjuvant fluorouracil chemotherapy in stage II cancer but a good response to immunotherapy in stage IV cancer. Therefore, determining MSI status in patients with colorectal cancer is important for identifying the appropriate treatment protocol. In the Pathology Artificial Intelligence Platform (PAIP) 2020 challenge, artificial intelligence researchers were invited to predict MSI status based on colorectal cancer slide images. Participants were required to perform two tasks. The primary task was to classify a given slide image as belonging to either the MSI-high or the microsatellite-stable group. The second task was tumor area segmentation to avoid ties with the main task. A total of 210 of the 495 participants enrolled in the challenge downloaded the images, and 23 teams submitted their final results. Seven teams from the top 10 participants agreed to disclose their algorithms, most of which were convolutional neural network-based deep learning models, such as EfficientNet and UNet. The top-ranked system achieved the highest F1 score (0.9231). This paper summarizes the various methods used in the PAIP 2020 challenge. This paper supports the effectiveness of digital pathology for identifying the relationship between colorectal cancer and the MSI characteristics.

Characterisation and validation of an in vitro transactivation assay based on the 22Rv1/MMTV_GR-KO cell line to detect human androgen receptor agonists and antagonists.

We describe the characterisation and validation of an androgen receptor (AR) transactivation assay for detection of AR agonists and antagonists using a stably transfected human prostate cancer cell line. This 22Rv1/mouse mammary tumour virus glucocorticoid knock-out cell line based AR transactivation assay was validated by criteria in Organisation for Economic Cooperation and Development Guidance Document 34 to determine if the assay performed equally well to the AR EcoScreen Assay included in Test Guideline for AR Transactivation (OECD TG 458). There was no Glucocorticoid Receptor (GR) crosstalk, and no changes in the AR DNA sequence in cells after the successful knock out of GR. Subsequently, the concordance of classifications of the 22 test chemicals was 100% in all laboratories. The AR agonistic and antagonistic inter-laboratory coefficients of variation based on log[10% effect for 10 nM DHT, PC10] and log[inhibitory response of 800 pM DHT by at 30%, IC30] from comprehensive tests were 2.75% and 2.44%, respectively. The AR agonist/antagonist test chemical classifications were consistent across AR EcoScreen ARTA assay data for 82/89%, and the balanced accuracy, sensitivity, and specificity were 83/90%, 88/100% and 78/80%, respectively. This assay was successfully validated and was approved for inclusion in TG 458 in 2020.

Quantification of bisphenols in Korean urine using online solid-phase extraction-high-performance liquid chromatography-tandem mass spectrometry.

Bisphenol A (BPA), an endocrine-disrupting chemical, has been used as a basic raw material for the production of polycarbonate plastics. As concern over the toxic effects of BPA grows, it is gradually being replaced in many consumer products with compounds such as bisphenol F (BPF) and bisphenol S (BPS). In this study, online solid-phase extraction-high-performance liquid chromatography-tandem mass spectrometry was used to analyze the urinary concentrations of BPA, BPF, and BPS in 2487 Korean urine samples collected between 2017 and 2018. The detection rates and geometric mean (GM) concentrations were as follows: BPA (82.1 %; 0.65 µg/L), BPF (11.1 %; not calculated), and BPS (63.6 %; 0.20 µg/L), respectively. The mean daily intake based on urinary BPA concentrations was 0.013 µg/kg bw/day (95th percentile, 0.089 µg/kg bw/day), which is lower than the tolerable daily intake. This is the first study observing urinary BPA, BPF, and BPS concentrations based on a nationally representative Korean population and could contribute to the evaluation of bisphenol analogue exposure levels in risk assessments.

LC-MS/MS analysis of BADGE, NOGEs, and their derivatives migrated from food and beverage metal cans.

A simple and novel method based on liquid chromatography-tandem mass spectrometry (LC-MS/MS) was developed and validated to determine the levels of 10 compounds (bisphenol A diglycidyl ether (BADGE), novolac-glycidyl ethers (NOGEs), and their related compounds) migrated from food and beverage cans into food simulants. Method validation showed acceptable linearity, precision, and accuracy. The detection limits ranged from 0.28 to 14.8 µg L-1, and the quantification limits ranged from 0.94 to 49.3 µg L-1. Water, 4% acetic acid, 50% ethanol, and n-heptane were employed as food simulants for the migration tests, and the developed LC-MS/MS method was applied to 104 epoxy-coated beverage and food metal cans. Only BADGE∙2H2O and BADGE were detected; the levels were below the specific migration limit. Based on the obtained migration results, the estimated daily intakes (EDIs) of BADGE∙2H2O and BADGE were calculated. Exposure assessments were conducted to compare the EDI with the tolerable daily intake (TDI), with a relatively low percentage of the TDI being reported. NOGE and its related compounds were not detected in the monitored cans. Long-term storage tests were also conducted at 60°C over 30 days. Only BADGE∙2H2O was detected in all food simulants, except for n-heptane, and the maximum amount detected was 114.6 µg L-1 in 50% ethanol.

Prevalence and antimicrobial resistance of Campylobacter jejuni and Campylobacter coli isolated from raw chicken meat and human stools in Korea.

Prevalence of Campylobacter in raw chicken meat and human stools and subsequent antibiotic resistance profiles of the pathogenic isolates obtained from 2000 through 2002 were investigated. Campylobacter jejuni and Campylobacter coli were isolated from 570 of the 923 raw chicken meat samples collected from traditional markets, large retail stores, or department stores in Korea, resulting in the isolation rate of 61.8%. A total of 579 Campylobacter isolates were obtained from raw chicken (36.3% for C. jejuni and 26.4% for C. coli) with the average population of 335.6 CFU/g. From 513 human stool samples, 15 isolates of Campylobacter were detected. Seasonal variation in the quantification of C. coli was not noticeable throughout the year, while the isolation rate of C. jejuni was the highest in September through October (840 CFU/g) followed by that of July through August and May through June in decreasing order, showing a significant seasonal effect (P < 0.05). Contamination of Campylobacter was more severe in raw chicken meat sold in traditional markets than in those sold in large retail stores and department stores. Prevalence of Campylobacter in raw chicken sold in traditional markets was significantly influenced by seasonal changes (P < 0.05), whereas the samples obtained from other places was less affected by the seasonal changes. Susceptibilities of the 594 chicken isolates to ciprofloxaxin, chloramphenicol, erythromycin, kanamycin, nalidixic acid, and tetracycline were determined by an E-test. Campylobacter isolates were the most resistant to nalidixic acid (91.4%) followed by ciprofloxaxin (87.9%), tetracycline (87.2%), kanamycin (30.6%), erythromycin (19.4%), and chloramphenicol (1.3%). Human isolates showed a similar resistance to the six antibiotics tested. The proportion of Campylobacter isolates with multidrug resistance to four or more antimicrobials obtained from 2000 through 2002 ranged from 28 to 43.5%, indicating that it could be a serious health-threatening factor. This study suggests that it is prudent to establish an effective National Monitoring Program in Korea for the prevention and control of Campylobacter spp.

Breakfast patterns are associated with metabolic syndrome in Korean adults.

The Korean diet, including breakfast, is becoming more Western, which could increase the risk of metabolic syndrome. Our aim was to assess whether breakfast patterns are associated with risk for metabolic syndrome in Korean adults. The study subjects (n = 371; 103 men, 268 women) were employees of Jaesang Hospital in Korea and their acquaintances, and all subjects were between 30 and 50 years old. The data collected from each subject included anthropometric measurements, three-day food intake, blood pressure (BP) and blood analyses. The three breakfast patterns identified by factor analysis were "Rice, Kimchi and Vegetables", "Potatoes, Fruits and Nuts" and "Eggs, Breads and Processed meat". The "Rice, Kimchi and Vegetables" pattern scores were positively correlated with systolic (SBP) and diastolic blood pressure (DBP) measurements in men (P < 0.05) and with serum triglyceride (TG) levels in women (P < 0.05). The "Eggs, Breads and Processed meat" pattern scores correlated positively with weight, body mass index (P < 0.05) and serum TGs (P < 0.01) in men. The "Potatoes, Fruits and Nuts" pattern was associated with lower risk of elevated BP (OR 0.49, 95% CI 0.28-0.88) and fasting glucose levels (OR 0.51, 95% CI 0.26-1.00). In contrast, the "Eggs, Breads and Processed meat" pattern was associated with increased risk of elevated TGs (OR 2.06, 95% CI 1.06-3.98). Our results indicate that reducing the consumption of eggs, western grains and processed meat while increasing fruit, nut and vegetable intake for breakfast could have beneficial effects on decreasing metabolic syndrome risk in Korean adults.

Skipping breakfast is associated with diet quality and metabolic syndrome risk factors of adults.

The aim of the present study was to assess the effects of skipping breakfast on diet quality and metabolic disease risk factors in healthy Korean adults. Subjects included 415 employees (118 men, 297 women; 30-50 years old) of Jaesang Hospital in Korea and their acquaintances. Data collected from each subject included anthropometric measurements, 3-day dietary intake, blood pressure, and blood analyses. The subjects were classified into three groups based on the number of days they skipped breakfast: 'Regular breakfast eater', 'Often breakfast eater', or 'Rare breakfast eater'. Participants in the 'Rare breakfast eater' group consumed less rice, potatoes, kimchi, vegetables, fish and shellfish, milk and dairy products, and sweets than did participants in the other two groups (P for trend < 0.05) and ate more cookies, cakes, and meat for dinner (P for trend < 0.05). Participants in the 'Rare breakfast eater' group consumed less daily energy, fat, dietary fiber, calcium, and potassium than did participants in the other groups (P for trend < 0.05). The percent energy from carbohydrates was lower and fat intake was higher in the 'Rare breakfast eater' group than in the other groups (P for trend < 0.01). When diets were compared using the Acceptable Macronutrient Distribution Range for Koreans, 59.1% of subjects in the 'Rare breakfast eater' group consumed more energy from fat compared with the other two groups (P < 0.005). According to the Estimated Average Requirements for Koreans, intake of selected nutrients was lower in the 'Rare breakfast eater' group than in the other two groups (P < 0.05). The risk of elevated serum triglycerides was decreased in the 'Rare breakfast eater' group (OR, 0.3 [0.1-1.0], P for trend = 0.0232). We conclude that eating breakfast regularly enhances diet quality, but may increase the risk of elevated serum triglycerides.