A deep learning model for predicting risks of crop pests and diseases from sequential environmental data.

Crop pests reduce productivity, so managing them through early detection and prevention is essential. Data from various modalities are being used to predict crop diseases by applying machine learning methodology. In particular, because growth environment data is relatively easy to obtain, many attempts are made to predict pests and diseases using it. In this paper, we propose a model that predicts diseases through previous growth environment information of crops, including air temperature, relative humidity, dew point, and CO2 concentration, using deep learning techniques. Using large-scale public data on crops of strawberry, pepper, grape, tomato, and paprika, we showed the model can predict the risk score of crop pests and diseases. It showed high predictive performance with an average AUROC of 0.917, and based on the predicted results, it can help prevent pests or post-processing. This environmental data-based crop disease prediction model and learning framework are expected to be universally applicable to various facilities and crops for disease/pest prevention.

Effect of regional medical disparities on complications in patients with hypertension: Cox's proportional hazard models.

Objective: Complications associated with hypertension can be alleviated by providing necessary medical services. However, there may be disparities in their provision depending on regional differences. Thus, this study aimed to examine the effects of regional healthcare disparities on complications in patients with hypertension in South Korea. Methods: Data from the National Health Insurance Service National Sample Cohort (2004-2019) were analyzed. The position value for the relative composite index was used to identify medically vulnerable regions. The diagnosis of hypertension within the region was also considered. The risk of complications associated with hypertension included cardiovascular, cerebrovascular, and kidney diseases. Cox proportional hazards models were used for statistical analysis. Results: A total of 246,490 patients were included in this study. Patients who lived in medically vulnerable regions and were diagnosed outside their residential area had a higher risk of complications than those living in non-vulnerable regions and were diagnosed outside the residential area (hazard ratio: 1.156, 95% confidence interval: 1.119-1.195). Conclusion: Patients living in medically vulnerable regions who were diagnosed outside their residential areas were more likely to have hypertension complications regardless of the type of complication. Necessary policies should be implemented to reduce regional healthcare disparities.

Longitudinal trajectories of sedation level and clinical outcomes in patients who are mechanically ventilated based on a group-based trajectory model: a prospective, multicentre, longitudinal and observational study in Korea.

OBJECTIVES: Changes in sedation levels over a long time in patients who are mechanically ventilated are unknown. Therefore, we investigated the long-term sedation levels of these patients by classifying them into different longitudinal patterns. DESIGN: This was a multicentre, prospective, longitudinal, and observational study. SETTING: Twenty intensive care units (ICUs) spanning several medical institutions in Korea. PARTICIPANTS: Patients who received mechanical ventilation and sedatives in ICU within 48 hours of admission between April 2020 and July 2021. PRIMARY AND SECONDARY OUTCOME MEASURES: The primary objective of this study was to identify the pattern of sedation practice. Additionally, we analysed the associations of trajectory groups with clinical outcomes as the secondary outcome. RESULTS: Sedation depth was monitored using Richmond Agitation-Sedation Scale (RASS). A group-based trajectory model was used to classify 631 patients into four trajectories based on sedation depth: persistent suboptimal (13.2%, RASS ≤ -3 throughout the first 30 days), delayed lightening (13.9%, RASS ≥ -2 after the first 15 days), early lightening (38.4%, RASS ≥ -2 after the first 7 days) and persistent optimal (34.6%, RASS ≥ -2 during the first 30 days). 'Persistent suboptimal' trajectory was associated with delayed extubation (HR: 0.23, 95% CI: 0.16 to 0.32, p<0.001), longer ICU stay (HR: 0.36, 95% CI: 0.26 to 0.51, p<0.001) and hospital mortality (HR: 13.62, 95% CI: 5.99 to 30.95, p<0.001) compared with 'persistent optimal'. The 'delayed lightening' and 'early lightening' trajectories showed lower extubation probability (HR: 0.30, 95% CI: 0.23 to 0.41, p<0.001; HR: 0.72, 95% CI: 0.59 to 0.87, p<0.001, respectively) and ICU discharge (HR: 0.44, 95% CI: 0.33 to 0.59, p<0.001 and HR: 0.80, 95% CI: 0.65 to 0.97, p=0.024) compared with 'persistently optimal'. CONCLUSIONS: Among the four trajectories, 'persistent suboptimal' trajectory was associated with higher mortality.

The Profile of Early Sedation Depth and Clinical Outcomes of Mechanically Ventilated Patients in Korea.

BACKGROUND: Current international guidelines recommend against deep sedation as it is associated with worse outcomes in the intensive care unit (ICU). However, in Korea the prevalence of deep sedation and its impact on patients in the ICU are not well known. METHODS: From April 2020 to July 2021, a multicenter, prospective, longitudinal, noninterventional cohort study was performed in 20 Korean ICUs. Sedation depth extent was divided into light and deep using a mean Richmond Agitation-Sedation Scale value within the first 48 hours. Propensity score matching was used to balance covariables; the outcomes were compared between the two groups. RESULTS: Overall, 631 patients (418 [66.2%] and 213 [33.8%] in the deep and light sedation groups, respectively) were included. Mortality rates were 14.1% and 8.4% in the deep and light sedation groups (P = 0.039), respectively. Kaplan-Meier estimates showed that time to extubation (P < 0.001), ICU length of stay (P = 0.005), and death (P = 0.041) differed between the groups. After adjusting for confounders, early deep sedation was only associated with delayed time to extubation (hazard ratio [HR], 0.66; 95% confidence interval [CI], 0.55-0.80; P < 0.001). In the matched cohort, deep sedation remained significantly associated with delayed time to extubation (HR, 0.68; 95% CI, 0.56-0.83; P < 0.001) but was not associated with ICU length of stay (HR, 0.94; 95% CI, 0.79-1.13; P = 0.500) and in-hospital mortality (HR, 1.19; 95% CI, 0.65-2.17; P = 0.582). CONCLUSION: In many Korean ICUs, early deep sedation was highly prevalent in mechanically ventilated patients and was associated with delayed extubation, but not prolonged ICU stay or in-hospital death.

Regional Differences in Medical Costs of Chronic Kidney Disease in the South Korean Population: Marginalized Two-Part Model.

BACKGROUND: There are regional gaps in the access to medical services for patients with chronic kidney disease (CKD), and it is necessary to reduce those gaps, including the gaps involving medical costs. OBJECTIVE: This study aimed to analyze regional differences in the medical costs associated with CKD in the South Korean population. METHODS: This longitudinal cohort study included participants randomly sampled from the National Health Insurance Service-National Sample Cohort of South Korea. To select those who were newly diagnosed with CKD, we excluded those who were diagnosed in 2002-2003 and 2018-2019. A total of 5903 patients with CKD were finally included. We used a marginalized two-part longitudinal model to assess total medical costs. RESULTS: Our cohort included 4775 (59.9%) men and 3191 (40.1%) women. Of these, 971 (12.2%) and 6995 (87.8%) lived in medically vulnerable and nonvulnerable regions, respectively. The postdiagnosis costs showed a significant difference between the regions (estimate: -0.0152, 95% confidence limit: -0.0171 to -0.0133). The difference in medical expenses between the vulnerable and nonvulnerable regions showed an increase each year after the diagnosis. CONCLUSIONS: Patients with CKD living in medically vulnerable regions are likely to have higher postdiagnostic medical expenses compared to those living in regions that are not medically vulnerable. Efforts to improve early diagnosis of CKD are needed. Relevant policies should be drafted to decrease the medical costs of patients with CKD disease living in medically deprived areas.

Detecting strawberry diseases and pest infections in the very early stage with an ensemble deep-learning model.

Detecting early signs of plant diseases and pests is important to preclude their progress and minimize the damages caused by them. Many methods are developed to catch signs of diseases and pests from plant images with deep learning techniques, however, detecting early signs is still challenging because of the lack of datasets to train subtle changes in plants. To solve these challenges, we built an automatic data acquisition system for the accumulation of a large dataset of plant images and trained an ensemble model to detect targeted plant diseases and pests. After obtaining 13,393 plant image data, our ensemble model shows a decent detection performance with an average of AUPRC 0.81. Also, this data acquisition and the detection process can be applied to other plant anomalies with the collection of additional data.

Difference in medical service use among Korean gastric cancer survivors according to regional healthcare vulnerabilities: a cohort study.

PURPOSE: This study investigated the relationship between medical service use and healthcare vulnerability, pre- and post-gastric cancer diagnosis. Differences between healthcare-vulnerable and healthcare-nonvulnerable regions identified inequities that require intervention. METHODS: This cohort study was done using the National Health Insurance claims data of patients diagnosed with gastric cancer between 2004 and 2013. The Position Value for Relative Comparison Index was used to determine whether the patients lived in a healthcare-vulnerable region. Medical service use was classified into annual outpatient treatment, hospitalization days, and emergency treatment. We used a generalized linear model to which the Poisson distribution was applied and compared regional differences in medical service use. RESULTS: A total of 1797 gastric cancer patients who had survived 5 years post-diagnosis were included in the study, of which 14.2% lived in healthcare-vulnerable regions. The patients in vulnerable regions surviving 5-7 years post-diagnosis had a higher number of outpatient visits than those in nonvulnerable regions. Furthermore, hospitalization days were lesser for patients in vulnerable regions who survived 6 years post-diagnosis than those in nonvulnerable regions; however, this number increased in the seventh year. CONCLUSIONS: Our results suggest that gastric cancer survivors living in healthcare-vulnerable regions have a higher probability of increased medical service use 5 years post-diagnosis compared with patients in nonvulnerable regions, which may significantly increase healthcare disparities over time. Therefore, in the future, additional research is needed to elucidate the causes of the disparities in healthcare use and the results of the differences in health outcomes.

Healthcare vulnerability disparities in pancreatic cancer treatment and mortality using the Korean National Sample Cohort: a retrospective cohort study.

BACKGROUND: The gap in treatment and health outcomes after diagnosis of pancreatic cancer is a major public health concern. We aimed to investigate the differences in the health outcomes and treatment of pancreatic cancer patients in healthcare vulnerable and non-vulnerable areas. METHODS: This retrospective cohort study evaluated data from the Korea National Health Insurance Corporation-National Sample Cohort from 2002 to 2019. The position value for relative comparison index was used to define healthcare vulnerable areas. Cox proportional hazard regression was used to estimate the risk of mortality in pancreatic cancer patients according to healthcare vulnerable areas, and multiple logistic regression was used to estimate the difference in treatment. RESULTS: Among 1,975 patients, 279 (14.1%) and 1,696 (85.9%) lived in the healthcare vulnerable and non-vulnerable areas, respectively. Compared with the non-vulnerable area, pancreatic cancer patients in the vulnerable area had a higher risk of death at 3 months (hazard ratio [HR]: 1.33, 95% confidence interval [CI] = 1.06-1.67) and 6 months (HR: 1.23, 95% CI = 1.03-1.48). In addition, patients with pancreatic cancer in the vulnerable area were less likely to receive treatment than patients in the non-vulnerable area (odds ratio [OR]: 0.70, 95% CI = 0.52-0.94). This trend was further emphasized for chemotherapy (OR: 0.68, 95% CI = 0.48-0.95). CONCLUSION: Patients with pancreatic cancer belonging to medically disadvantaged areas receive less treatment and have a higher risk of death. This may be a result of the late diagnosis of pancreatic cancer among these patients.