We aimed to use deep learning to detect tuberculosis in chest radiographs in annual workers’ health examination data and compare the performances of convolutional neural networks (CNNs) based on images only (I-CNN) and CNNs including demographic variables (D-CNN). The I-CNN and D-CNN models were trained on 1000 chest X-ray images, both positive and negative, for tuberculosis. Feature extraction was conducted using VGG19, InceptionV3, ResNet50, DenseNet121, and InceptionResNetV2. Age, weight, height, and gender were recorded as demographic variables. The area under the receiver operating characteristic (ROC) curve (AUC) was calculated for model comparison. The AUC values of the D-CNN models were greater than that of I-CNN. The AUC values for VGG19 increased by 0.0144 (0.957 to 0.9714) in the training set, and by 0.0138 (0.9075 to 0.9213) in the test set (both p < 0.05). The D-CNN models show greater sensitivity than I-CNN models (0.815 vs. 0.775, respectively) at the same cut-off point for the same specificity of 0.962. The sensitivity of D-CNN does not attenuate as much as that of I-CNN, even when specificity is increased by cut-off points. Conclusion: Our results indicate that machine learning can facilitate the detection of tuberculosis in chest X-rays, and demographic factors can improve this process.

Background Discrimination is associated with depressive symptoms and other negative health effects, but little is known about the mental health risks of workplace gender discrimination. We aimed to investigate the association of workplace gender discrimination and depressive symptoms among employed women in South Korea. Methods The 6 th wave (2016) survey datasets of the Korean Longitudinal Survey of Women and Family (KLoWF) were analyzed for 2,339 respondents who are identified as wage workers. Depressive symptoms were evaluated by the short-form (10-item) Center for Epidemiological Studies-Depression scale. Association of workplace gender discrimination and depressive symptoms was assessed using multivariate logistic regression, adjusted for potential confounding variables including age, income satisfaction, education level, marital status, and currently diagnosed disease. We then measured the age effect using age stratification multivariate logistic regression model. Results Women who experienced gender discrimination at workplace had higher odds of depressive symptoms regardless of the type of the discrimination including hiring, promotion, work assignments, paid wages, and firing. These associations were consistent in younger women below 40 years of age in regard to hiring, promotion, paid wages and firing, whereas inconsistent among older women above 40 years of age. Limitations We did not investigate the effect of workplace gender discrimination on depressive symptoms in a longitudinal manner. Conclusions Workplace gender discrimination was found to be significantly associated with depressive symptoms after adjustment for socio-demographic factors. Further, women under 40 years of age were especially vulnerable to workplace gender discrimination.

BackgroundThere is a lack of statistical analysis investigating the relationship between sleep problems and commute time in Korea. We aimed to analyze the association between representative health symptoms, sleep disturbances, and commute time according to working hours in Korea.MethodsThe 4th Korean Working Conditions Survey data were used for analysis, and unpaid family workers and workers who work fewer than three days in a week were excluded. Commute time, working hours, and sleep hours were assessed using self-reported questionnaires. Odds ratios (ORs) with 95% confidence intervals (CIs) for sleep problems were calculated using a multivariate logistic regression model with ≤10 min commute time as the reference group.ResultsAmong a total of 28,804 workers (men = 14,945, women = 13,859), 2.6% of men and 3.2% of women experienced sleep problems. In both sexes, long commute time (51–60 minutes and >60 minutes) showed an increased OR [men, 2.03 (CI = 1.32–3.13) and 2.05 (CI = 1.33–3.17); women, 1.58 (CI = 1.05–2.39) and 1.63 (CI = 1.06–2.50), respectively]. In stratification analysis of working hours, long commute time (51–60 and > 60 minutes) showed an increased OR in men working >40 hours/week [2.08 (CI = 1.16–3.71) and 1.92 (CI = 1.08–3.41), respectively]. Furthermore, long commute time (41–50, 51–60, and >60 minutes) showed an increased OR in women working >40 hours/week [2.40 (CI = 1.27–4.55), 2.28 (CI = 1.25–4.16), and 2.19 (CI = 1.17–4.16), respectively]. Moreover, commute time >60 minutes showed an increased OR in women working ≤40 hours/week [1.96 (CI = 1.06–3.62)].ConclusionThis large cross-sectional study highlights that long commute time is related to sleep problems in both sexes. Shorter commute times and decreased working hours are needed to prevent sleep problems in workers.

No abstract

previous studies have suggested that in addition to respiratory system cancers, exposure to external airborne agents (EAAs) may also affect the risk of digestive tract cancer. However, previous epidemiological studies have been limited. To clarify this relationship, we conducted a Workers' Korea National Health Insurance Service cohort study. The EAA exposure group comprised participants who had ever visited a hospital as an inpatient for 'lung diseases due to external agents'. The reference population comprised men from the general working population. The EAA exposure group and reference group included a total of 98,666 and 79,959,286 person-years, respectively. Age-adjusted standardized incident rates (SIRs) with 95% confidence intervals (95%CI) were calculated for each 5-year age stratum. The SIR (95% CI) of EAA exposure was 1.30 (1.19-1.38) for all digestive tract cancers. The highest risk associated with eAA exposure was observed for oral cancer, followed by esophageal and stomach cancers [SIRs (95%CI): 3.96 (3.02-4.78), 3.47(2.60-4.25), and 1.34(1.17-1.47), respectively.] These statistically significant associations did not be attenuated in a subgroup analysis using logistic regression adjusted for age, smoking and alcohol consumption. Our findings suggest that EAA exposure should address risk reduction of both digestive tract and respiratory system cancers. Previous studies of the human health effects of external airborne agent (EAA) which were included aerosol, gas, vapor, mist, fume, dust, or smoke exposure have focused on the respiratory system 1-3 , given the assumption that such exposure primarily causes respiratory system disease. Although EAA exposure mainly occurs via the respiratory system, humans face other possible sites of exposure, namely the digestive system. The digestive system may be exposed to EAAs via several inhalation and ingestion mechanisms. First, agents filtered in the nose or exhaled from the upper respiratory tract may be swallowed 4. Second,, reduced sphincter tone around the esophageal orifice could directly allow the accidental swallowing of EAAs 5. Third, food, skin, or clothing may be contaminated by EAAs 4. Finally, an intuitive approach reveals that the origins of the gastrointestinal tract and respiratory system occur in shared structures, such as the oral cavity and pharynx. In summary, these mechanisms identify the gastrointestinal tract as a potential site of EAA exposure, with the potential for inflammatory, immunologic, or even oncogenic responses similar to those observed in the respiratory system 6. To date, little is known about the relationship between EAA exposure and the risk of digestive tract cancers. Notably, a previous well-designed cohort study found that esophageal and stomach cancer were more closely correlated with EAA exposure when compared with lung cancer, although the authors did not control for alcohol consumption 7. Nevertheless, these results encouraged subsequent epidemiological studies. Although later research also demonstrated the effects of E...