Deep Neural Network-Based Concentration Model for Oak Pollen Allergy Warning in South Korea

PURPOSE: Oak is the dominant tree species in Korea. Oak pollen has the highest sensitivity rate among all allergenic tree species in Korea. A deep neural network (DNN)-based estimation model was developed to determine the concentration of oak pollen and overcome the shortcomings of conventional regression models. METHODS: The DNN model proposed in this study utilized weather factors as the input and provided pollen concentrations as the output. Weather and pollen concentration data were used from 2007 to 2016 obtained from the Korea Meteorological Administration pollen observation network. Because it is difficult to prevent over-fitting and underestimation by using a DNN model alone, we developed a bootstrap aggregating-type ensemble model. Each of the 30 ensemble members was trained with random sampling at a fixed rate according to the pollen risk grade. To verify the effectiveness of the proposed model, we compared its performance with those of models of regression and support vector regression (SVR) under the same conditions, with respect to the prediction of pollen concentrations, risk levels, and season length. RESULTS: The mean absolute percentage error in the estimated pollen concentrations was 11.18%, 10.37%, and 5.04% for the regression, SVR and DNN models, respectively. The start of the pollen season was estimated to be 20, 22, and 6 days earlier than that predicted by the regression, SVR and DNN models, respectively. Similarly, the end of the pollen season was estimated to be 33, 20, and 9 days later that predicted by the regression, SVR and DNN models, respectively. CONCLUSIONS: Overall, the DNN model performed better than the other models. However, the prediction of peak pollen concentrations needs improvement. Improved observation quality with optimization of the DNN model will resolve this issue.

Allergenic Pollen Calendar in Korea Based on Probability Distribution Models and Up-to-Date Observations

PURPOSE: The pollen calendar is the simplest forecasting method for pollen concentrations. As pollen concentrations are liable to seasonal variations due to alterations in climate and land-use, it is necessary to update the pollen calendar using recent data. To attenuate the impact of considerable temporal and spatial variability in pollen concentrations on the pollen calendar, it is essential to employ a new methodology for its creation.METHODS: A pollen calendar was produced in Korea using data from recent observations, and a new method for creating the calendar was proposed, considering both risk levels and temporal resolution of pollen concentrations. A probability distribution was used for smoothing concentrations and determining risk levels. Airborne pollen grains were collected between 2007 and 2017 at 8 stations; 13 allergenic pollens, including those of alder, Japanese cedar, birch, hazelnut, oak, elm, pine, ginkgo, chestnut, grasses, ragweed, mugwort and Japanese hop, were identified from the collected grains.RESULTS: The concentrations of each pollen depend on locations and seasons due to large variability in species distribution and their environmental condition. In the descending order of concentration, pine, oak and Japanese hop pollens were found to be the most common in Korea. The pollen concentrations were high in spring and autumn, and those of oak and Japanese hop were probably the most common cause of allergy symptoms in spring and autumn, respectively. High Japanese cedar pollen counts were observed in Jeju, while moderate concentrations were in Jeonju, Gwangju and Busan.CONCLUSIONS: A new methodology for the creation of a pollen calendar was developed to attenuate the impact of large temporal and spatial variability in pollen concentrations. This revised calendar should be available to the public and allergic patients to prevent aggravation of pollen allergy.

Revision of threshold levels for evoking pollinosis to oak, pine, Japanese hop, and ragweed in the metropolitan area Seoul, Korea

Purpose@#The threshold levels for symptom development of pollinosis vary among studies and countries. This study aimed to determine currently used threshold levels for it. @*Methods@#Oak, pine, Japanese hop, and ragweed pollen samples were collected daily for 8 years from the Seoul and Guri areas. A total of 792 subjects with allergy to these pollens were recruited. The symptom index (SI) was assessed through telephone interviews and allergy questionnaires, and data were analyzed using decision tree. @*Results@#The risk index for oak pollen allergy was “mild” when the pollen count was 0–2 grains/m3 , “moderate” when it was 3–11 grains/m3 , “severe” when it was 12–28 grains/m3 , and “dangerous” when it was ≥ 29 grains/m3 . The risk level for pine pollen allergy was “mild” when the pollen count was 0–4 grains/m 3 , “moderate” when it was 5–42 grains/m3 , “severe” when it was 43–66 grains/m3 , and “dangerous” when it was ≥ 67 grains/m3 . For Japanese hop pollen allergy, the risk level was “mild” when the pollen count was 0–8 grains/m3 , “moderate” when it was 9–10 grains/m3 , “severe” when it was 11–19 grains/m3 , and “dangerous” when it was ≥ 20 grains/m3 . Finally, for ragweed, the risk level was “mild” when the pollen count was 0–1 grains/m3 , “moderate” when it was 2–6 grains/m3 , “severe” when it was 7–33 grains/m3 , and “dangerous” when it was ≥ 34 grains/m3 . @*Conclusions@#Revising the threshold levels for the risk index for pollen allergies may be useful for developing pollen prediction models for patients with pollen allergies in Korea.

Chamber and Field Studies demonstrate Differential Amb a 1 Contents in Common Ragweed Depending on CO₂ Levels

Although atmospheric carbon dioxide (CO₂) has no apparent direct effect on human health, it does have direct effects on plants. The present study evaluated the influence of increased CO₂ levels on the concentration of allergens from common ragweed pollen by setting up a chamber study to model future air conditions and a field study to evaluate current air conditions. For the chamber study, we established 20 ragweed plants in an open-top chamber under different CO₂ levels (380–400, 500–520, 600–620, and 1,000–1,100 parts per million [ppm]). For the field study, we established ragweed plants in rural (Pocheon, Gyeonggi-do; mean CO₂ 320±54.8 ppm) and urban (Gangnam, Seoul; mean CO₂ 440±78.5 ppm) locations. Seeds of the common ragweed (Ambrosia artemisiifolia) were obtained from Daejin University. The Amb a 1 protein content of pollen extracts was quantified using a double sandwich enzyme-linked immunosorbent assay. In our chamber study, the median concentration of Amb a 1 in pollen increased with increasing in CO₂ concentration (1.88 ng/µg in 380–400 ppm CO₂; 3.14 ng/µg in 500–520 ppm CO₂; 4.44 ng/µg in 600–620 ppm CO₂; and 5.36 ng/µg in 1,000–1,100 ppm CO₂). In our field study, we found no significantly different concentration of Amb a 1 between the pollen extracts at the Pocheon (mean±standard deviation, 1.63±0.3 ng/µg pollen in 320±54.8 ppm CO₂) and the Gangnam (2.04±0.7 ng/µg pollen in CO₂ in 440±78.5 ppm CO₂) locations, although the concentration of Amb a 1 was increased in the Gangnam than in the Pocheon locations. Our results suggest that future increases in CO₂ levels to more than 600 ppm will significantly elevate the Amb a 1 content in common ragweeds, although the current different CO₂ levels do not cause differences in the Amb a 1 content of ragweed pollen.

Evaluation of the association between pollen count and the outbreak of allergic disease

PURPOSE: This study focused on the evaluation of the relation between pollen concentration and the outbreak of allergic disease (symptom index), and this outcome would be necessary to upgrade risk grade for the pollen forecasting system. METHODS: Airborne particles carrying allergens, such as pollen, were collected daily at the Seoul and Guri area by using 7-day Burkard samplers for 6 years. A total of 596 Subjects were recruited from Hanyang University Seoul Hospital (n=144 for spring, n=139 for autumn), and Hanyang University Guri Hospital (n=157 for spring, n=156 for autumn). Symptom index was evaluated and recorded by phone calling to study subjects daily or asking questionnaire when they visit outpatient clinic every week. Statistical analysis of data was performed by using correlation coefficients and regression models with time series graph. RESULTS: Two peak seasons of pollen concentration were May and September in Korea. In skin prick tests, the sensitization rate to ragweed pollen was gradually increased in children. In the same period, sensitization rates to airborne pollen, especially oak, birch for spring, and Japanese hop for autumn were increased annually. There was a significantly relationship between symptom index of allergic patients and allergic pollen concentrations in this study. Especially symptom index was significantly correlated to the concentration of oak pollen of day 1 in spring and to the concentration of Japanese hop pollen of day 0 in autumn. CONCLUSION: Sensitization rates to pollens increased annually. There is a significant relationship between allergy symptom index and pollen concentration. There remains to confirm the Korean own risk grade of pollen allergy.

A Six-Year Study on the Changes in Airborne Pollen Counts and Skin Positivity Rates in Korea: 2008-2013

PURPOSE: The occurrence of pollen allergy is subject to exposure to pollen, which shows regional and temporal variations. We evaluated the changes in pollen counts and skin positivity rates for 6 years, and explored the correlation between their annual rates of change. MATERIALS AND METHODS: We assessed the number of pollen grains collected in Seoul, and retrospectively reviewed the results of 4442 skin-prick tests conducted at the Severance Hospital Allergy-Asthma Clinic from January 1, 2008 to December 31, 2013. RESULTS: For 6 years, the mean monthly total pollen count showed two peaks, one in May and the other in September. Pollen count for grasses also showed the same trend. The pollen counts for trees, grasses, and weeds changed annually, but the changes were not significant. The annual skin positivity rates in response to pollen from grasses and weeds increased significantly over the 6 years. Among trees, the skin positivity rates in response to pollen from walnut, popular, elm, and alder significantly increased over the 6 years. Further, there was a significant correlation between the annual rate of change in pollen count and the rate of change in skin positivity rate for oak and hop Japanese. CONCLUSION: The pollen counts and skin positivity rates should be monitored, as they have changed annually. Oak and hop Japanese, which showed a significant correlation with the annual rate of change in pollen count and the rate of change in skin positivity rate over the 6 years may be considered the major allergens in Korea.

ERRATUM: Affiliation Correction. Evaluation of the association of vegetation of allergenic plants and pollinosis with meteorological changes

According to the author's request, in this paper, the eighth author's (Bong-Seong Kim) affiliation should be corrected.

Evaluation of the association of vegetation of allergenic plants and pollinosis with meteorological changes

PURPOSE: There are a number of reports suggesting that widespread propagation of weeds and high concentration of weed pollen have been contributed to climate change. We investigated the interrelationship between allergenic pollen concentration, allergic symptom and meteorological factor. METHODS: We collected data of pollen concentration and meteorological factors in 7 stations nationwide during between 1998 and 2012. We recruited total 297 allergic patients sensitized to weed pollens from each station, conducted a survey about allergic symptom, and calculated symptom index. We surveyed the vegetation area of ragweed and Japanese hop. Based on these data, we performed the long-term trend analysis (X11-ARIMA, autoregressive integrated moving average) on regional pollen concentration, and correlation analysis to investigate the interrelation between weed pollen concentration, allery symptom index and meteorological factor. We have also done regression analysis on vegetation area and maximal pollen concentration. RESULTS: Long-term trend analysis showed the increasing trend of pllen concentration in Seoul. Weed pollen concentration, allergy symptom index and each meteorological factor were not correlated significantly. Regression analysis revealed that increase of weed vegetation area results in increase of weed pollen concentration. Through this regression equation, we estimated the vegetation area that can product pollen concentration triggering allergenic risk. CONCLUSION: Meteorological factors, pollen concentration and allergic symptoms should be consistently assessed and the relationship between each factor should be analyzed, considering climate change. It is necessary to verify the equation for pollen estimation by vegetation area and set up a policy for vegetation control focused on the reduction of allergenic pollen.

The Revised Edition of Korean Calendar for Allergenic Pollens

The old calendar of pollens did not reflect current pollen distribution and concentrations that can be influenced by changes of weather and environment of each region in South Korea. A new pollen calendar of allergenic pollens was made based on the data on pollen concentrations obtained in eight regions nationwide between 1997 and 2009. The distribution of pollen was assessed every day at 8 areas (Seoul, Guri, Busan, Daegu, Jeonju, Kwangju, Kangneung, and Jeju) for 12 years between July 1, 1997 and June 30, 2009. Pollens were collected by using Burkard 7-day sampler (Burkard Manufacturing Co Ltd, UK). Pollens which were stained with Calberla's fuchsin staining solution were identified and counted. Pine became the highest pollen in May, and the pollen concentrations of oak and birch also became high. Ragweed appeared in the middle of August and showed the highest pollen concentration in the middles of September. Japanese hop showed a high concentration between the middle of August and the end of September, and mugwort appeared in the middles of August and its concentration increased up until early September. In Kangneung, birch appeared earlier, pine showed a higher pollen concentration than in the other areas. In Daegu, Oriental thuja and alder produced a large concentration of pollens. Pine produced a large concentration of pollens between the middle of April and the end of May. Weeds showed higher concentrations in September and mugwort appeared earlier than ragweed. In Busan the time of flowering is relatively early, and alder and Oriental thuja appeared earliest among all areas. In Kwangju, Oriental thuja and hazelnut appeared in early February. Japanese cedar showed the highest pollen concentration in March in Jeju. In conclusion, update information on pollen calendar in South Korea should be provided for allergic patients through the website to manage and prevent the pollinosis.

Changes in Sensitization Rate to Weed Allergens in Children with Increased Weeds Pollen Counts in Seoul Metropolitan Area

The prevalence of allergic diseases in children has increased for several decades. We evaluated the correlation between pollen count of weeds and their sensitization rate in Seoul, 1997-2009. Airborne particles carrying allergens were collected daily from 3 stations around Seoul. Skin prick tests to pollen were performed on children with allergic diseases. Ragweed pollen gradually increased between 1999 and 2005, decreased after 2005 and plateaued until 2009 (peak counts, 67 in 2003, 145 in 2005 and 83 grains/m3/day in 2007). Japanese hop pollen increased between 2002 and 2009 (peak counts, 212 in 2006 and 492 grains/m3/day in 2009). Sensitization rates to weed pollen, especially ragweed and Japanese hop in children with allergic diseases, increased annually (ragweed, 2.2% in 2000 and 2.8% in 2002; Japanese hop, 1.4% in 2000 and 1.9% in 2002). The age for sensitization to pollen gradually became younger since 2000 (4 to 6 yr of age, 3.5% in 1997 and 6.2% in 2009; 7 to 9 yr of age, 4.2% in 1997 and 6.4% in 2009). In conclusion, sensitization rates for weed pollens increase in Korean children given increasing pollen counts of ragweed and Japanese hop.

Changes in Sensitization Rate to Weed Allergens in Children with Increased Weeds Pollen Counts in Seoul Metropolitan Area

The prevalence of allergic diseases in children has increased for several decades. We evaluated the correlation between pollen count of weeds and their sensitization rate in Seoul, 1997-2009. Airborne particles carrying allergens were collected daily from 3 stations around Seoul. Skin prick tests to pollen were performed on children with allergic diseases. Ragweed pollen gradually increased between 1999 and 2005, decreased after 2005 and plateaued until 2009 (peak counts, 67 in 2003, 145 in 2005 and 83 grains/m3/day in 2007). Japanese hop pollen increased between 2002 and 2009 (peak counts, 212 in 2006 and 492 grains/m3/day in 2009). Sensitization rates to weed pollen, especially ragweed and Japanese hop in children with allergic diseases, increased annually (ragweed, 2.2% in 2000 and 2.8% in 2002; Japanese hop, 1.4% in 2000 and 1.9% in 2002). The age for sensitization to pollen gradually became younger since 2000 (4 to 6 yr of age, 3.5% in 1997 and 6.2% in 2009; 7 to 9 yr of age, 4.2% in 1997 and 6.4% in 2009). In conclusion, sensitization rates for weed pollens increase in Korean children given increasing pollen counts of ragweed and Japanese hop.