Enhancing automated strabismus classification with limited data: Data augmentation using StyleGAN2-ADA.

In this study, we propose a generative data augmentation technique to overcome the challenges of severely limited data when designing a deep learning-based automated strabismus diagnosis system. We implement a generative model based on the StyleGAN2-ADA model for system design and assess strabismus classification performance using two classifiers. We evaluate the capability of our proposed method against traditional data augmentation techniques and confirm a substantial enhancement in performance. Furthermore, we conduct experiments to explore the relationship between the diagnosis agreement among ophthalmologists and the generation performance of the generative model. Beyond FID, we validate the generative samples on the classifier to establish their practicality. Through these experiments, we demonstrate that the generative model-based data augmentation improves overall quantitative performance in scenarios of extreme data scarcity and effectively mitigates overfitting issues during deep learning model training.

Relationship between workplace violence and work-related depression/anxiety, separating the types of perpetrators: a cross-sectional study using data from the fourth and fifth Korean Working Conditions Surveys (KWCS).

Background: Work is an inseparable element of a person's life, and violence in the workplace has various effects on individual workers and companies. While most studies have focused on specific industries, very few studies have investigated the influence of workplace violence by co-workers. Therefore, this study aimed to evaluate the association between workplace violence and work-related depression/anxiety in various occupations by differentiating the perpetrators of violence as co-workers and clients. Methods: This study was conducted based on data from the 4th and 5th Korean Working Conditions Surveys (KWCS). The experience of workplace violence was classified in terms of the perpetrator: workplace violence by co-workers and that by clients. Work-related depression and anxiety were assessed using questions about health problems related to depression and anxiety and whether the problems were related to work. Descriptive statistics, χ2 tests, and multiple logistic regression analyses were performed using the SPSS 26.0. Results: After adjusting for sociodemographic characteristics (age, education, income, subjective health status) and occupational characteristics (occupation, weekly working hours, type of employment, size of workplace, and shift work), male workers with experience of workplace violence by co-workers were found to be at a higher risk of work-related depression/anxiety (odds ratio [OR], 11.52; 95% confidence interval [CI], 8.65-15.36). The same was confirmed for female workers (OR, 10.89; 95% CI, 7.90-15.02). Conclusions: Employees who experienced workplace violence from co-workers were found to be more vulnerable to work-related depression/anxiety. Continuous contact between the victim and the perpetrator may occur, and the possibility of a secondary assault can frighten the victim. Appropriate prevention and intervention measures that focus on the perpetrators of violence are needed.

Comparing risk of depression between day and night/shift workers using the PHQ-9: a study utilizing the 2014, 2016, and 2018 Korea National Health and Nutrition Examination Survey data.

BACKGROUND: In today's work scenario, the number of shift workers, including those in night shifts, is increasing. Shift work can adversely affect workers' health in the long run, but studies on the relationship between shift work and depression have shown inconsistent results. This study aimed to determine whether the number of night/shift workers at risk of depression, as predicted by the Patient Health Questionnaire-9 (PHQ-9), is higher than that of day workers. METHODS: This study was conducted based on data from the 6th and 7th Korea National Health and Nutrition Examination Survey, 2014, 2016, and 2018. Work schedules were classified into 2 types: day work and night/shift work. This study used the PHQ-9, a self-reported depression screening test, to identify workers at risk of depression. Statistical analysis was performed using SPSS 26.0, and descriptive statistics, χ2 test, and logistic regression analysis were employed. RESULTS: After adjusting for age, educational level, working hours per week, and income, men engaging in night/shift work were at a higher risk of depression (odds ratio [OR]: 1.407, 95% confidence interval [CI]: 0.937-2.113). The same was confirmed for women (OR: 1.564, 95% CI: 1.176-2.081). CONCLUSIONS: Our results showed that the OR for those engaged in night/shift work with a PHQ-9 score of 10 or more increased. Considering the large volume of psychiatric history and symptoms in Korea, additional research is needed. Additionally, further discussion on ways to provide realistic help to night/shift workers is warranted.

Association between sleep duration and impaired fasting glucose according to work type in non-regular workers: data from the first and second year (2016, 2017) of the 7th Korean National Health and Nutrition Examination (KNHANE) (a cross-sectional study).

BACKGROUND: We aimed to find the relationship between sleep duration and impaired fasting glucose according to working type in non-regular workers using the 2016 and 2017 Korean National Health And Nutrition Examination (KNHANE, 7th revision). METHOD: In the 1st and 2nd year (2016, 2017) of the 7th KNHANE, 16,277 people participated. Minors were excluded because this study was intended for individuals aged 19 years and older. As this study was based on wage workers, unemployment, self-employed workers, employers, unpaid family workers, and those who have insufficient answers such as unknown or no response were excluded. Regular workers were excluded because this study was intended for non-regular workers. Finally, a total of 2,168 people were included in the survey, except those who had been diagnosed with diabetes, had a fasting blood glucose level of 126 mg/dL or higher, or taking hypoglycemic agents or receiving insulin injections. To find the relationship between sleep duration and impaired fasting glucose according to work type in non-regular workers, multiple logistic regression analysis was performed by adjusting the general and occupational characteristics after stratification according to work type. All statistical analyses were performed using the SPSS software (version 26.0; SPSS Inc., Chicago, IL, USA). RESULTS: In the case of insufficient sleep duration in irregular female workers, the odds ratio (OR) of impaired fasting glucose was statistically insignificant, but in the case of insufficient sleep duration in irregular male workers who have shift work, the odds ratio (OR) of impaired fasting glucose was significantly higher than that of sufficient sleep duration (Model 1, OR: 3.05, 95% confidence interval [CI]: 1.18-7.90; Model 2, OR: 2.81, 95% CI: 1.08-7.29). CONCLUSIONS: Our findings demonstrate that insufficient sleep duration was associated with an increase in fasting blood glucose levels in non-regular male workers working shifts. This means that non-regular workers are in desperate need for adequate sleep and health care. We hope that our study will help improve the health of non-regular workers and more systematic and prospective follow-up studies will be conducted to further improve the health of non-regular workers.