An otoscope is traditionally used to examine the eardrum and ear canal.A diagnosis of otitis media(OM)relies on the experience of clinicians.If an examiner lacks experience,the examination may be difficult and time-co...An otoscope is traditionally used to examine the eardrum and ear canal.A diagnosis of otitis media(OM)relies on the experience of clinicians.If an examiner lacks experience,the examination may be difficult and time-consuming.This paper presents an ear disease classification method using middle ear images based on a convolutional neural network(CNN).Especially the segmentation and classification networks are used to classify an otoscopic image into six classes:normal,acute otitis media(AOM),otitis media with effusion(OME),chronic otitis media(COM),congenital cholesteatoma(CC)and traumatic perforations(TMPs).The Mask R-CNN is utilized for the segmentation network to extract the region of interest(ROI)from otoscopic images.The extracted ROIs are used as guiding features for the classification.The classification is based on transfer learning with an ensemble of two CNN classifiers:EfficientNetB0 and Inception-V3.The proposed model was trained with a 5-fold cross-validation technique.The proposed method was evaluated and achieved a classification accuracy of 97.29%.展开更多
Adequate oxygen in red blood cells carrying through the body to the heart and brain is important to maintain life.For those patients requiring blood,blood transfusion is a common procedure in which donated blood or bl...Adequate oxygen in red blood cells carrying through the body to the heart and brain is important to maintain life.For those patients requiring blood,blood transfusion is a common procedure in which donated blood or blood components are given through an intravenous line.However,detecting the need for blood transfusion is time-consuming and sometimes not easily diagnosed,such as internal bleeding.This study considered physiological signals such as electrocardiogram(ECG),photoplethysmogram(PPG),blood pressure,oxygen saturation(SpO2),and respiration,and proposed the machine learning model to detect the need for blood transfusion accurately.For the model,this study extracted 14 features from the physiological signals and used an ensemble approach combining extreme gradient boosting and random forest.The model was evaluated by a stratified five-fold crossvalidation:the detection accuracy and area under the receiver operating characteristics were 92.7%and 0.977,respectively.展开更多
Both cigarette and e-cigarette use cause respiratory tract damage and related health outcomes,and potentially increase the risk of coronavirus disease 2019(COVID-19)-related symptoms[1].Since the COVID-19 pandemic,loc...Both cigarette and e-cigarette use cause respiratory tract damage and related health outcomes,and potentially increase the risk of coronavirus disease 2019(COVID-19)-related symptoms[1].Since the COVID-19 pandemic,local and central governments have legally mandated wearing masks indoors and outdoors[2].展开更多
The coronavirus disease 2019(COVID-19)pandemic has raised concerns about the mental health and social well-being of youth,including its potential to increase or exacerbate substance use behaviors[1].Among adolescents,...The coronavirus disease 2019(COVID-19)pandemic has raised concerns about the mental health and social well-being of youth,including its potential to increase or exacerbate substance use behaviors[1].Among adolescents,the COVID-19pandemic has resulted in limited face-to-face school contact and thus missed milestones in preventing alcohol and substance use.展开更多
For the classification problem in practice,one of the challenging issues is to obtain enough labeled data for training.Moreover,even if such labeled data has been sufficiently accumulated,most datasets often exhibit l...For the classification problem in practice,one of the challenging issues is to obtain enough labeled data for training.Moreover,even if such labeled data has been sufficiently accumulated,most datasets often exhibit long-tailed distribution with heavy class imbalance,which results in a biased model towards a majority class.To alleviate such class imbalance,semisupervised learning methods using additional unlabeled data have been considered.However,as a matter of course,the accuracy is much lower than that from supervised learning.In this study,under the assumption that additional unlabeled data is available,we propose the iterative semi-supervised learning algorithms,which iteratively correct the labeling of the extra unlabeled data based on softmax probabilities.The results show that the proposed algorithms provide the accuracy as high as that from the supervised learning.To validate the proposed algorithms,we tested on the two scenarios:with the balanced unlabeled dataset and with the imbalanced unlabeled dataset.Under both scenarios,our proposed semi-supervised learning algorithms provided higher accuracy than previous state-of-the-arts.Code is available at https://github.com/HeewonChung92/iterative-semi-learning.展开更多
Background Although smoking is classified as a risk factor for severe COVID-19 outcomes,there is a scarcity of studies on prevalence of smoking during the COVID-19 pandemic.Thus,this study aims to analyze the trends o...Background Although smoking is classified as a risk factor for severe COVID-19 outcomes,there is a scarcity of studies on prevalence of smoking during the COVID-19 pandemic.Thus,this study aims to analyze the trends of prevalence of smoking in adolescents over the COVID-19 pandemic period.Methods The present study used data from middle to high school adolescents between 2005 and 2021 who participated in the Korea Youth Risk Behavior Web-based Survey(KYRBS).We evaluated the smoking prevalence(ever or daily)by year groups and estimated the slope in smoking prevalence before and during the pandemic.Results A total of 1,137,823 adolescents participated in the study[mean age,15.04 years[95%confidence interval(CI)15.03-15.06];and male,52.4%(95%CI 51.7-53.1)].The prevalence of ever smokers was 27.7%(95%CI 27.3-28.1)between 2005 and 2008 but decreased to 9.8%(95%CI 9.3-10.3)in 2021.A consistent trend was found in daily smokers,as the estimates decreased from 5.4%(95%CI 5.2-5.6)between 2005 and 2008 to 2.3%(95%CI 2.1-2.5)in 2021.However,the downward slope in the overall prevalence of ever smokers and daily smokers became less pronounced in the COVID-19 pandemic period than in the pre-pandemic period.In the subgroup with substance use,the decreasing slope in daily smokers was significantly more pronounced during the pandemic than during the pre-pandemic period.Conclusions The proportion of ever smokers and daily smokers showed a less pronounced decreasing trend during the pandemic.The findings of our study provide an overall understanding of the pandemic's impact on smoking prevalence in adolescents.展开更多
Background Although previous studies have provided data on early pandemic periods of alcohol and substance use in adolescents,more adequate studies are needed to predict the trends of alcohol and substance use during ...Background Although previous studies have provided data on early pandemic periods of alcohol and substance use in adolescents,more adequate studies are needed to predict the trends of alcohol and substance use during recent periods,including the mid-pandemic period.This study investigated the changes in alcohol and substance use,except tobacco use,throughout the pre-,early-,and mid-pandemic periods in adolescents using a nationwide serial cross-sectional survey from South Korea.Methods Data on 1,109,776 Korean adolescents aged 13–18 years from 2005 to 2021 were obtained in a survey operated by the Korea Disease Control and Prevention Agency.We evaluated adolescents’alcohol and substance consumption prevalence and compared the slope of alcohol and substance prevalence before and during the COVID-19 pandemic to see the trend changes.We define the pre-COVID-19 period as consisting of four groups of consecutive years(2005–2008,2009–2012,2013–2015,and 2016–2019).The COVID-19 pandemic period is composed of 2020(early-pandemic era)and 2021(midpandemic era).Results More than a million adolescents successfully met the inclusion criteria.The weighted prevalence of current alcohol use was 26.8%[95%confidence interval(CI)26.4–27.1]from 2005 to 2008 and 10.5%(95%CI 10.1–11.0)in 2020 and 2021.The weighted prevalence of substance use was 1.1%(95%CI 1.1–1.2)from 2005 to 2008 and 0.7%(95%CI 0.6–0.7)between 2020 and 2021.From 2005 to 2021,the overall trend of use of both alcohol and drugs was found to decrease,but the decline has slowed since COVID-19 epidemic(current alcohol use:βdiff 0.167;95%CI 0.150–0.184;substance use:βdiff 0.152;95%CI 0.110–0.194).The changes in the slope of current alcohol and substance use showed a consistent slowdown with regard to sex,grade,residence area,and smoking status from 2005 to 2021.Conclusion The overall prevalence of alcohol consumption and substance use among over one million Korean adolescents from the early and mid-stage(2020–2021)of the COVID-19 pandemic showed a slower decline than expected given the increase during the prepandemic period(2005–2019).展开更多
基金This study was supported by a Basic Science Research Program through the National Research Foundation of Korea(NRF)funded by the Ministry of Science,ICT&Future Planning NRF-2020R1A2C1014829the Soonchunhyang University Research Fund.
文摘An otoscope is traditionally used to examine the eardrum and ear canal.A diagnosis of otitis media(OM)relies on the experience of clinicians.If an examiner lacks experience,the examination may be difficult and time-consuming.This paper presents an ear disease classification method using middle ear images based on a convolutional neural network(CNN).Especially the segmentation and classification networks are used to classify an otoscopic image into six classes:normal,acute otitis media(AOM),otitis media with effusion(OME),chronic otitis media(COM),congenital cholesteatoma(CC)and traumatic perforations(TMPs).The Mask R-CNN is utilized for the segmentation network to extract the region of interest(ROI)from otoscopic images.The extracted ROIs are used as guiding features for the classification.The classification is based on transfer learning with an ensemble of two CNN classifiers:EfficientNetB0 and Inception-V3.The proposed model was trained with a 5-fold cross-validation technique.The proposed method was evaluated and achieved a classification accuracy of 97.29%.
基金This work was supported by the Korea Medical Device Development Fund from the Korean government(the Ministry of Science and ICTMinistry of Trade,Indus-try and Energy+2 种基金Ministry of Health and Welfareand Ministry of Food and Drug Safety)(KMDF_PR_20200901_0095)the Soonchunhyang University Research Fund.
文摘Adequate oxygen in red blood cells carrying through the body to the heart and brain is important to maintain life.For those patients requiring blood,blood transfusion is a common procedure in which donated blood or blood components are given through an intravenous line.However,detecting the need for blood transfusion is time-consuming and sometimes not easily diagnosed,such as internal bleeding.This study considered physiological signals such as electrocardiogram(ECG),photoplethysmogram(PPG),blood pressure,oxygen saturation(SpO2),and respiration,and proposed the machine learning model to detect the need for blood transfusion accurately.For the model,this study extracted 14 features from the physiological signals and used an ensemble approach combining extreme gradient boosting and random forest.The model was evaluated by a stratified five-fold crossvalidation:the detection accuracy and area under the receiver operating characteristics were 92.7%and 0.977,respectively.
基金the Korea Health Technology R&D Project through the Korea Health Industry Development Institute(KHIDI)funded by the Ministry of Health&Welfare,Republic of Korea(grant number:HV22C0233)+2 种基金the National Research Foundation of Korea(NRF)(MSITRS-2023-00248157)a grant(21153MFDS601)from the Ministry of Food and Drug Safety in 2023
文摘Both cigarette and e-cigarette use cause respiratory tract damage and related health outcomes,and potentially increase the risk of coronavirus disease 2019(COVID-19)-related symptoms[1].Since the COVID-19 pandemic,local and central governments have legally mandated wearing masks indoors and outdoors[2].
基金supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute(KHIDI)funded by the Ministry of Health&WelfareRepublic of Korea[grant number:HV22C0233]。
文摘The coronavirus disease 2019(COVID-19)pandemic has raised concerns about the mental health and social well-being of youth,including its potential to increase or exacerbate substance use behaviors[1].Among adolescents,the COVID-19pandemic has resulted in limited face-to-face school contact and thus missed milestones in preventing alcohol and substance use.
基金This work was supported by the National Research Foundation of Korea(No.2020R1A2C1014829)by the Korea Medical Device Development Fund grant,which is funded by the Government of the Republic of Korea Korea government(the Ministry of Science and ICT+2 种基金the Ministry of Trade,Industry and Energythe Ministry of Health and Welfareand the Ministry of Food and Drug Safety)(grant KMDF_PR_20200901_0095).
文摘For the classification problem in practice,one of the challenging issues is to obtain enough labeled data for training.Moreover,even if such labeled data has been sufficiently accumulated,most datasets often exhibit long-tailed distribution with heavy class imbalance,which results in a biased model towards a majority class.To alleviate such class imbalance,semisupervised learning methods using additional unlabeled data have been considered.However,as a matter of course,the accuracy is much lower than that from supervised learning.In this study,under the assumption that additional unlabeled data is available,we propose the iterative semi-supervised learning algorithms,which iteratively correct the labeling of the extra unlabeled data based on softmax probabilities.The results show that the proposed algorithms provide the accuracy as high as that from the supervised learning.To validate the proposed algorithms,we tested on the two scenarios:with the balanced unlabeled dataset and with the imbalanced unlabeled dataset.Under both scenarios,our proposed semi-supervised learning algorithms provided higher accuracy than previous state-of-the-arts.Code is available at https://github.com/HeewonChung92/iterative-semi-learning.
基金supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute(KHIDI),funded by the Ministry of Health&Welfare,Republic of Korea(grant number:HV22C0233)the National Research Foundation of Korea(NRF)grant funded by the Korea government(NRF2021R1I1A2059735)。
文摘Background Although smoking is classified as a risk factor for severe COVID-19 outcomes,there is a scarcity of studies on prevalence of smoking during the COVID-19 pandemic.Thus,this study aims to analyze the trends of prevalence of smoking in adolescents over the COVID-19 pandemic period.Methods The present study used data from middle to high school adolescents between 2005 and 2021 who participated in the Korea Youth Risk Behavior Web-based Survey(KYRBS).We evaluated the smoking prevalence(ever or daily)by year groups and estimated the slope in smoking prevalence before and during the pandemic.Results A total of 1,137,823 adolescents participated in the study[mean age,15.04 years[95%confidence interval(CI)15.03-15.06];and male,52.4%(95%CI 51.7-53.1)].The prevalence of ever smokers was 27.7%(95%CI 27.3-28.1)between 2005 and 2008 but decreased to 9.8%(95%CI 9.3-10.3)in 2021.A consistent trend was found in daily smokers,as the estimates decreased from 5.4%(95%CI 5.2-5.6)between 2005 and 2008 to 2.3%(95%CI 2.1-2.5)in 2021.However,the downward slope in the overall prevalence of ever smokers and daily smokers became less pronounced in the COVID-19 pandemic period than in the pre-pandemic period.In the subgroup with substance use,the decreasing slope in daily smokers was significantly more pronounced during the pandemic than during the pre-pandemic period.Conclusions The proportion of ever smokers and daily smokers showed a less pronounced decreasing trend during the pandemic.The findings of our study provide an overall understanding of the pandemic's impact on smoking prevalence in adolescents.
基金supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute(KHIDI)funded by the Ministry of Health&Welfare,Republic of Korea(grant number:HV22C0233 and grant number:HI22C1976)The funders had no role in study design,data collection,data analysis,data interpretation,or writing of the report.
文摘Background Although previous studies have provided data on early pandemic periods of alcohol and substance use in adolescents,more adequate studies are needed to predict the trends of alcohol and substance use during recent periods,including the mid-pandemic period.This study investigated the changes in alcohol and substance use,except tobacco use,throughout the pre-,early-,and mid-pandemic periods in adolescents using a nationwide serial cross-sectional survey from South Korea.Methods Data on 1,109,776 Korean adolescents aged 13–18 years from 2005 to 2021 were obtained in a survey operated by the Korea Disease Control and Prevention Agency.We evaluated adolescents’alcohol and substance consumption prevalence and compared the slope of alcohol and substance prevalence before and during the COVID-19 pandemic to see the trend changes.We define the pre-COVID-19 period as consisting of four groups of consecutive years(2005–2008,2009–2012,2013–2015,and 2016–2019).The COVID-19 pandemic period is composed of 2020(early-pandemic era)and 2021(midpandemic era).Results More than a million adolescents successfully met the inclusion criteria.The weighted prevalence of current alcohol use was 26.8%[95%confidence interval(CI)26.4–27.1]from 2005 to 2008 and 10.5%(95%CI 10.1–11.0)in 2020 and 2021.The weighted prevalence of substance use was 1.1%(95%CI 1.1–1.2)from 2005 to 2008 and 0.7%(95%CI 0.6–0.7)between 2020 and 2021.From 2005 to 2021,the overall trend of use of both alcohol and drugs was found to decrease,but the decline has slowed since COVID-19 epidemic(current alcohol use:βdiff 0.167;95%CI 0.150–0.184;substance use:βdiff 0.152;95%CI 0.110–0.194).The changes in the slope of current alcohol and substance use showed a consistent slowdown with regard to sex,grade,residence area,and smoking status from 2005 to 2021.Conclusion The overall prevalence of alcohol consumption and substance use among over one million Korean adolescents from the early and mid-stage(2020–2021)of the COVID-19 pandemic showed a slower decline than expected given the increase during the prepandemic period(2005–2019).