Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Kyulhee Han has a total of 42 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    An ensemble approach improves the prediction of the COVID-19 pandemic in South Korea

    by Kyulhee Han, Catherine Apio, Hanbyul Song, Bogyeom Lee, Xuwen Hu, Jiwon Park, Liu Zhe, Taewan Goo, Taesung Park

    Every infectious disease outbreak exhibits specific spread dynamics that need to be identified based on the transmission patterns of a given disease outbreak [1]. With every epidemic and pandemic, the biggest challenge is the perception of the outbreak by the population: governments focus on the intervention measures to be taken, while the general population fears […]


  • Open Access

    Forecasting the spread of COVID-19 based on policy, vaccination, and Omicron data

    by Kyulhee Han, Bogyeom Lee, Doeun Lee, Gyujin Heo, Jooha Oh, Seoyoung Lee, Catherine Apio, Taesung Park

    The COVID-19 pandemic caused by the novel SARS-COV-2 virus poses a great risk to the world. During the COVID-19 pandemic, observing and forecasting several important indicators of the epidemic (like new confirmed cases, new cases in intensive care unit, and new deaths for each day) helped prepare the appropriate response (e.g., creating additional intensive care unit beds, and implementing strict interventions). Various predictive models and predictor variables have been used to forecast these indicators. However, the impact of prediction models and predictor variables on forecasting performance has not been systematically well analyzed. Here, we compared the forecasting performance using a linear mixed model in terms of prediction models (mathematical, statistical, and AI/machine learning models) and predictor variables (vaccination rate, stringency index, and Omicron variant rate) for seven selected countries with the highest vaccination rates. We decided on our best models based on the Bayesian Information Criterion (BIC) and analyzed the significance of each predictor. Simple models were preferred. The selection of the best prediction models and the use of Omicron variant rate were considered essential in improving prediction accuracies. For the test data period before Omicron variant emergence, the selection of the best models was the most significant factor in improving prediction accuracy. For the test period after Omicron emergence, Omicron variant rate use was considered essential in deciding forecasting accuracy. For prediction models, ARIMA, lightGBM, and TSGLM generally performed well in both test periods. Linear mixed models with country as a random effect has proven that the choice of prediction models and the use of Omicron data was significant in determining forecasting accuracies for the highly vaccinated countries. Relatively simple models, fit with either prediction model or Omicron data, produced best results in enhancing forecasting accuracies with test data.


  • Open Access

    Development of New Stringency Indices for Nonpharmacological Social Distancing Policies Implemented in Korea During the COVID-19 Pandemic: Random Forest Approach

    by Catherine Apio, Kyulhee Han, Doeun Lee, Bogyeom Lee, Taesung Park

    Background: In the absence of an effective treatment method or vaccine, the outbreak of the COVID-19 pandemic elicited a wide range of unprecedented restriction policies aimed at mitigating and suppressing the spread of the SARS-CoV-2 virus. These policies and their Stringency Index (SI) of more than 160 countries were systematically recorded in the Oxford COVID-19 Government Response Tracker (OxCGRT) data set. The SI is a summary measure of the overall strictness of these policies. However, the OxCGRT SI may not fully reflect the stringency levels of the restriction policies implemented in Korea. Korea implemented 33 COVID-19 restriction policies targeting 4 areas: public facilities, public events, social gatherings, and religious gatherings. Objective: This study aims to develop new Korea Stringency Indices (KSIs) that reflect the stringency levels of Korea’s restriction policies better and to determine which government-implemented policies were most effective in managing the COVID-19 pandemic in Korea. Methods: The random forest method was used to calculate the new KSIs using feature importance values and determine their effectiveness in managing daily COVID-19 confirmed cases. Five analysis periods were considered, including November 01, 2020, to January 20, 2021 (Period 1), January 20, 2021, to June 27, 2021 (Period 2), November 01, 2020, to June 27, 2021 (Period 3), June 27, 2021, to November 01, 2021 (Period 4), and November 01, 2021, to April 24, 2022 (Period 5). Results: Among the KSIs, public facilities in period 4, public events in period 2, religious gatherings in periods 1 and 3, and social gatherings in period 5 had the highest importance. Among the public facilities, policies associated with operation hour restrictions in cinemas, restaurants, PC rooms, indoor sports facilities, karaoke, coffee shops, night entertainment facilities, and baths or saunas had the highest importance across all analysis periods. Strong positive correlations were observed between daily confirmed cases and public facilities, religious gatherings, and public events in period 1 of the pandemic. From then, weaker and negative correlations were observed in the remaining analysis periods. The comparison with the OxCGRT SI showed that the SI had a relatively lower feature importance and correlation with daily confirmed cases than the proposed KSIs, making KSIs more effective than SI. Conclusions: Restriction policies targeting public facilities were the most effective among the policies analyzed. In addition, different periods call for the enforcement of different policies given their effectiveness varies during the pandemic. Trial Registration:


  • Open Access

    An analysis of the waning effect of COVID-19 vaccinations

    by Bogyeom Lee, Hanbyul Song, Catherine Apio, Kyulhee Han, Jiwon Park, Zhe Liu, Hu Xuwen, Taesung Park

    Vaccine development is one of the key efforts to control the spread of coronavirus disease 2019 (COVID-19). However, it has become apparent that the immunity acquired through vaccination is not permanent, known as the waning effect. Therefore, monitoring the proportion of the population with immunity is essential to improve the forecasting of future waves of the pandemic. Despite this, the impact of the waning effect on forecasting accuracies has not been extensively studied. We proposed a method for the estimation of the effective immunity (EI) rate which represents the waning effect by integrating the second and booster doses of COVID-19 vaccines. The EI rate, with different periods to the onset of the waning effect, was incorporated into three statistical models and two machine learning models. Stringency Index, omicron variant BA.5 rate (BA.5 rate), booster shot rate (BSR), and the EI rate were used as covariates and the best covariate combination was selected using prediction error. Among the prediction results, Generalized Additive Model showed the best improvement (decreasing 86% test error) with the EI rate. Furthermore, we confirmed that South Korea’s decision to recommend booster shots after 90 days is reasonable since the waning effect onsets 90 days after the last dose of vaccine which improves the prediction of confirmed cases and deaths. Substituting BSR with EI rate in statistical models not only results in better predictions but also makes it possible to forecast a potential wave and help the local community react proactively to a rapid increase in confirmed cases.


  • Open Access

    Development of New Stringency Indices for Nonpharmacological Social Distancing Policies Implemented in Korea During the COVID-19 Pandemic: Random Forest Approach (Preprint)

    by Catherine Apio, Kyulhee Han, Doeun Lee, Bogyeom Lee, Taesung Park

    BACKGROUND In the absence of an effective treatment method or vaccine, the outbreak of the COVID-19 pandemic elicited a wide range of unprecedented restriction policies aimed at mitigating and suppressing the spread of the SARS-CoV-2 virus. These policies and their Stringency Index (SI) of more than 160 countries were systematically recorded in the Oxford COVID-19 Government Response Tracker (OxCGRT) data set. The SI is a summary measure of the overall strictness of these policies. However, the OxCGRT SI may not fully reflect the stringency levels of the restriction policies implemented in Korea. Korea implemented 33 COVID-19 restriction policies targeting 4 areas: public facilities, public events, social gatherings, and religious gatherings. OBJECTIVE This study aims to develop new Korea Stringency Indices (KSIs) that reflect the stringency levels of Korea’s restriction policies better and to determine which government-implemented policies were most effective in managing the COVID-19 pandemic in Korea. METHODS The random forest method was used to calculate the new KSIs using feature importance values and determine their effectiveness in managing daily COVID-19 confirmed cases. Five analysis periods were considered, including November 01, 2020, to January 20, 2021 (Period 1), January 20, 2021, to June 27, 2021 (Period 2), November 01, 2020, to June 27, 2021 (Period 3), June 27, 2021, to November 01, 2021 (Period 4), and November 01, 2021, to April 24, 2022 (Period 5). RESULTS Among the KSIs, public facilities in period 4, public events in period 2, religious gatherings in periods 1 and 3, and social gatherings in period 5 had the highest importance. Among the public facilities, policies associated with operation hour restrictions in cinemas, restaurants, PC rooms, indoor sports facilities, karaoke, coffee shops, night entertainment facilities, and baths or saunas had the highest importance across all analysis periods. Strong positive correlations were observed between daily confirmed cases and public facilities, religious gatherings, and public events in period 1 of the pandemic. From then, weaker and negative correlations were observed in the remaining analysis periods. The comparison with the OxCGRT SI showed that the SI had a relatively lower feature importance and correlation with daily confirmed cases than the proposed KSIs, making KSIs more effective than SI. CONCLUSIONS Restriction policies targeting public facilities were the most effective among the policies analyzed. In addition, different periods call for the enforcement of different policies given their effectiveness varies during the pandemic. CLINICALTRIAL

  • Open Access

    Microbiome markers of pancreatic cancer based on bacteria-derived extracellular vesicles acquired from blood samples: A retrospective propensity score matching analysis

    by Jae Ri Kim, Kyulhee Han, Youngmin Han, Nayeon Kang, Tae-Seop Shin, Hyeon Ju Park, Hongbeom Kim, Wooil Kwon, Seungyeoun Lee, Yoon-Keun Kim, et al.

    Jae Ri KIM, Kyulhee HAN, Youngmin HAN, Nayeon KANG, Tae-Seop SHIN, Hyeon Ju PARK, Hongbeom KIM, Wooil KWON, Seungyeoun LEE, Yoon-Keun KIM, Taesung PARK, Jin-Young JANG*. Ann Hepatobiliary Pancreat Surg 2022;26:S55. https://doi.org/10.14701/ahbps.2022S1.BR-OP-5

  • Open Access

    Forecasting of the COVID-19 pandemic situation of Korea

    by Taewan Goo, Catherine Apio, Gyujin Heo, Doeun Lee, Jong Hyeok Lee, Jisun Lim, Kyulhee Han, Taesung Park

    For the novel coronavirus disease 2019 (COVID-19), predictive modeling, in the literature, uses broadly susceptible exposed infected recoverd (SEIR)/SIR, agent-based, curve-fitting models. Governments and legislative bodies rely on insights from prediction models to suggest new policies and to assess the effectiveness of enforced policies. Therefore, access to accurate outbreak prediction models is essential to obtain insights into the likely spread and consequences of infectious diseases. The objective of this study is to predict the future COVID-19 situation of Korea. Here, we employed 5 models for this analysis; SEIR, local linear regression (LLR), negative binomial (NB) regression, segment Poisson, deep-learning based long short-term memory models (LSTM) and tree based gradient boosting machine (GBM). After prediction, model performance comparison was evelauated using relative mean squared errors (RMSE) for two sets of train (January 20, 2020‒December 31, 2020 and January 20, 2020‒January 31, 2021) and testing data (January 1, 2021‒February 28, 2021 and February 1, 2021‒February 28, 2021) . Except for segmented Poisson model, the other models predicted a decline in the daily confirmed cases in the country for the coming future. RMSE values' comparison showed that LLR, GBM, SEIR, NB, and LSTM respectively, performed well in the forecasting of the pandemic situation of the country. A good understanding of the epidemic dynamics would greatly enhance the control and prevention of COVID-19 and other infectious diseases. Therefore, with increasing daily confirmed cases since this year, these results could help in the pandemic response by informing decisions about planning, resource allocation, and decision concerning social distancing policies.


  • Open Access

    Microbiome Markers of Pancreatic Cancer Based on Bacteria-Derived Extracellular Vesicles Acquired from Blood Samples: A Retrospective Propensity Score Matching Analysis

    by Jae Ri Kim, Kyulhee Han, Youngmin Han, Nayeon Kang, Tae-Seop Shin, Hyeon Ju Park, Hongbeom Kim, Wooil Kwon, Seungyeoun Lee, Yoon-Keun Kim, et al.

    Novel biomarkers for early diagnosis of pancreatic cancer (PC) are necessary to improve prognosis. We aimed to discover candidate biomarkers by identifying compositional differences of microbiome between patients with PC (n = 38) and healthy controls (n = 52), using microbial extracellular vesicles (EVs) acquired from blood samples. Composition analysis was performed using 16S rRNA gene analysis and bacteria-derived EVs. Statistically significant differences in microbial compositions were used to construct PC prediction models after propensity score matching analysis to reduce other possible biases. Between-group differences in microbial compositions were identified at the phylum and genus levels. At the phylum level, three species (Verrucomicrobia, Deferribacteres, and Bacteroidetes) were more abundant and one species (Actinobacteria) was less abundant in PC patients. At the genus level, four species (Stenotrophomonas, Sphingomonas, Propionibacterium, and Corynebacterium) were less abundant and six species (Ruminococcaceae UCG-014, Lachnospiraceae NK4A136 group, Akkermansia, Turicibacter, Ruminiclostridium, and Lachnospiraceae UCG-001) were more abundant in PC patients. Using the best combination of these microbiome markers, we constructed a PC prediction model that yielded a high area under the receiver operating characteristic curve (0.966 and 1.000, at the phylum and genus level, respectively). These microbiome markers, which altered microbial compositions, are therefore candidate biomarkers for early diagnosis of PC.


  • Open Access

    Statistical Estimation of Effects of Implemented Government Policies on COVID-19 Situation in South Korea

    by Gyujin Heo, Catherine Apio, Kyulhee Han, Taewan Goo, Hye Won Chung, Taehyun Kim, Hakyong Kim, Yeonghyeon Ko, Doeun Lee, Jisun Lim, et al.

    Since the outbreak of novel SARS-COV-2, each country has implemented diverse policies to mitigate and suppress the spread of the virus. However, no systematic evaluation of these policies in their alleviation of the pandemic has been done. We investigate the impact of five indices derived from 12 policies in the Oxford COVID-19 Government Response Tracker dataset and the Korean government’s index, which is the social distancing level implemented by the Korean government in response to the changing pandemic situation. We employed segmented Poisson model for this analysis. In conclusion, health and the Korean government indices are most consistently effective (with negative coefficients), while the restriction and stringency indexes are mainly effective with lagging (1~10 days), as intuitively daily confirmed cases of a given day is affected by the policies implemented days before, which shows that a period of time is required before the impact of some policies can be observed. The health index demonstrates the importance of public information campaign, testing policy and contact tracing, while the government index shows the importance of social distancing guidelines in mitigating the spread of the virus. These results imply the important roles of these polices in mitigation of the spread of COVID-19 disease.