전염병의 경로 추적 및 예측을 위한 통합 정보 시스템 구현

김은경 ( Eungyeong Kim ),이석 ( Seok Lee ),변영태 ( Young Tae Byun ),이혁재 ( Hyuk-jae Lee ),이택진 ( Taikjin Lee ) 한국인터넷정보학회 2013 인터넷정보학회논문지 Vol.14 No.5

세계적으로 전파력과 병원성이 높은 신종인플루엔자, 조류독감 등과 같은 전염병이 증가하고 있다. 전염병이란 특정 병원체(pathogen)로 인하여 발생하는 질병으로 감염된 사람으로부터 감수성이 있는 숙주(사람)에게 감염되는 질환을 의미한다. 전염병의 병원체는 세균, 스피로헤타, 리케차, 바이러스, 진균, 기생충 등이 있으며, 호흡기계 질환, 위장관 질환, 간질환, 급성 열성 질환 등을 일으킨다. 전파 방법은 식품이나 식수, 곤충 매개, 호흡에 의한 병원체의 흡입, 다른 사람과의 접촉 등 다양한 경로를 통해 발생한다. 전 세계의 대부분 국가들은 전염병의 전파를 예측하고 대비하기 위해서 수학적 모델을 사용하고 있다. 하지만 과거와 달리 현대 사회는 지상과 지하 교통수단의 발달로 전염병의 전파 속도가 매우 복잡하고 빨라졌기 때문에 우리는 이를 예방하기 위한 대책 마련의 시간이 부족하다. 그러므로 전염병의 확산을 막기 위해서는 전염병의 전파 경로를 예측할 수 있는 시스템이 필요하다. 우리는 이러한 문제를 해결하기 위해서 전염병의 실시간 감시 및 관리를 위한 전염병의 감염 경로 추적 및 예측이 가능한 통합정보 시스템을 구현하였다. 이 논문에서는 전염병의 전파경로 예측에 관한 부분을 다루며, 이 시스템은 기존의 수학적 모델인Susceptible-Infectious-Recovered (SIR) 모델을 기반으로 하였다. 이 모델의 특징은 교통수단인 버스, 기차, 승용차, 비행기를 포함시킴으로써, 도시내 뿐만 아니라 도시간의 교통수단을 이용한 이동으로 사람간의 접촉을 표현할 수 있다. 그리고 한국의 지리적 특성에 맞도록 실제 자료를 수정하였기 때문에 한국의 현실을 잘 반영할 수 있다. 또한 백신은 시간에 따라서 투여 지역과 양을 조절할 수 있기 때문에 사용자가 시뮬레이션을 통해서 어느 시점에서 어느 지역에 우선적으로 투여할지 백신을 컨트롤할 수 있다. 시뮬레이션은 몇가지 가정과 시나리오를 기반으로 한다. 그리고 통계청의 자료를 이용해서 인구 이동이 많은 주요 5개 도시인서울, 인천국제공항, 강릉, 평창, 원주를 선정했다. 상기 도시들은 네트워크로 연결되어있으며 4가지의 교통수단들만 이용하여 전파된다고 가정하였다. 교통량은 국가통계포털에서 일일 교통량 자료를 입수하였으며, 각도시의 인구수는 통계청에서 통계자료를 입수하였다. 그리고 질병관리본부에서는 신종인플루엔자 A의 자료를 입수하였으며, 항공포털시스템에서는 항공 통계자료를 입수하였다. 이처럼 일일 교통량, 인구 통계, 신종인플루엔자 A 그리고 항공 통계자료는 한국의 지리적 특성에 맞도록 수정하여 현실에 가까운 가정과 시나리오를 바탕으로 하였다. 시뮬레이션은 신종인플루엔자 A가 인천공항에 발생하였을 때, 백신이 투여되지 않은 경우, 서울과 평창에 각각 백신이 투여된 경우의 3가지 시나리오에 대해서, 감염자가 피크인 날짜와 I (infectious)의 비율을 비교하였다. 그 결과 백신이 투여되지 않은 경우, 감염자가 피크인 날짜는 교통량이 가장 많은 서울에서 37일로 가장 빠르고, 교통량이 가장 적은 평창에서 43일로 가장 느렸다. I의 비율은 서울에서 가장 높았고, 평창에서 가장 낮았다. 서울에 백신이 투여된 경우, 감염자가 피크인 날짜는 서울이 37일로 가장 빨랐으며, 평창은 43일로 가장 느렸다. 그리고 I의 비율은 강릉에서 가장 높으며, 평창에서 가장 낮았다. 평창에 백신을 투여한 경우, 감염자가 피크인 날짜는 37일로 서울이 가장 빠르고 평창은 43일로 가장 느렸다. I의 비율은 강릉에서 가장 높았고, 평창에서는 가장 낮았다. 이 결과로부터 신종인플루엔자 A가 발생하면 각 도시는 교통량에 의해 영향을 받아 확산된다는 것을 확인할 수 있다. 따라서 전염병 발생시 전파 경로는 각 도시의 교통량에 따라서 달라지므로, 교통량의 분석을 통해서 전염병의 전파 경로를 추적하고 예측함으로써 전염병에 대한 대책이 가능할 것이다. The incidence of globally infectious and pathogenic diseases such as H1N1 (swine flu) and Avian Influenza (AI) has recently increased. An infectious disease is a pathogen-caused disease, which can be passed from the infected person to the susceptible host. Pathogens of infectious diseases, which are bacillus, spirochaeta, rickettsia, virus, fungus, and parasite, etc., cause various symptoms such as respiratory disease, gastrointestinal disease, liver disease, and acute febrile illness. They can be spread through various means such as food, water, insect, breathing and contact with other persons. Recently, most countries around the world use a mathematical model to predict and prepare for the spread of infectious diseases. In a modern society, however, infectious diseases are spread in a fast and complicated manner because of rapid development of transportation (both ground and underground). Therefore, we do not have enough time to predict the fast spreading and complicated infectious diseases. Therefore, new system, which can prevent the spread of infectious diseases by predicting its pathway, needs to be developed. In this study, to solve this kind of problem, an integrated monitoring system, which can track and predict the pathway of infectious diseases for its realtime monitoring and control, is developed. This system is implemented based on the conventional mathematical model called by `Susceptible-Infectious-Recovered (SIR) Model.` The proposed model has characteristics that both inter- and intra-city modes of transportation to express interpersonal contact (i.e., migration flow) are considered. They include the means of transportation such as bus, train, car and airplane. Also, modified real data according to the geographical characteristics of Korea are employed to reflect realistic circumstances of possible disease spreading in Korea. We can predict where and when vaccination needs to be performed by parameters control in this model. The simulation includes several assumptions and scenarios. Using the data of Statistics Korea, five major cities, which are assumed to have the most population migration have been chosen; Seoul, Incheon (Incheon International Airport), Gangneung, Pyeongchang and Wonju. It was assumed that the cities were connected in one network, and infectious disease was spread through denoted transportation methods only. In terms of traffic volume, daily traffic volume was obtained from Korean Statistical Information Service (KOSIS). In addition, the population of each city was acquired from Statistics Korea. Moreover, data on H1N1 (swine flu) were provided by Korea Centers for Disease Control and Prevention, and air transport statistics were obtained from Aeronautical Information Portal System. As mentioned above, daily traffic volume, population statistics, H1N1 (swine flu) and air transport statistics data have been adjusted in consideration of the current conditions in Korea and several realistic assumptions and scenarios. Three scenarios (occurrence of H1N1 in Incheon International Airport, not-vaccinated in all cities and vaccinated in Seoul and Pyeongchang respectively) were simulated, and the number of days taken for the number of the infected to reach its peak and proportion of Infectious (I) were compared. According to the simulation, the number of days was the fastest in Seoul with 37 days and the slowest in Pyeongchang with 43 days when vaccination was not considered. In terms of the proportion of I, Seoul was the highest while Pyeongchang was the lowest. When they were vaccinated in Seoul, the number of days taken for the number of the infected to reach at its peak was the fastest in Seoul with 37 days and the slowest in Pyeongchang with 43 days. In terms of the proportion of I, Gangneung was the highest while Pyeongchang was the lowest. When they were vaccinated in Pyeongchang, the number of days was the fastest in Seoul with 37 days and the slowest in Pyeongchang with 43 days. In terms of the proportion of I, Gangneung was the highest while Pyeongchang was the lowest. Based on the results above, it has been confirmed that H1N1, upon the first occurrence, is proportionally spread by the traffic volume in each city. Because the infection pathway is different by the traffic volume in each city, therefore, it is possible to come up with a preventive measurement against infectious disease by tracking and predicting its pathway through the analysis of traffic volume.

볼륨-플로우 그래프 기반 폐질환 분류를 위한앙상블 딥러닝 모델

김재운,김홍준 한국지능시스템학회 2024 한국지능시스템학회논문지 Vol.34 No.1

Chronic Obstructive Pulmonary Disease (COPD) is a respiratory disease characterized bychronic airway obstruction. COPD often progresses to a severe stage, since thereare few noticeable symptoms in the early stages. Also regression equations involingvarious factors such as race, gender, height, and weight to determine whether ornot there is pulmonary disease is complex and needs to be updated periodically. Therefore, there is a demand for a system that can easily analze the presence orabsence of the pulmonary disease, even for non-experts. In this paper, aCNN-based flow volume loops classification model using ensemble learning andappropriate data pre-processing algorithms was proposed and validated to diagnosepulmonary disease in the early stages. The ensemble model was organized by fourCNN models based on VGG16, VGG19, Resnet50, and MobileNet and used transferlearning and fine-tuning for each pre-trained model. Specifically, to overcome asmall amount of data, several data augmentation techniques that took into accountthe characteristics of flow volume loops were used, and soft voting was employedfor the ensemble model. The proposed ensemble model not only could diagnose thepresence or absence of pulmonary disease but could also classify into a total of fourcategories: normal, restrictive, obstructive, and combined pulmonary diseases. As aresult of the experiment, the performance of the proposed ensemble model showedan accuracy of 90.91%, precision of 91.11%, and recall of 90.91%. 만성 폐쇄성 폐질환은 만성적인 기도 폐쇄를 특징으로 하는 호흡기 질환이다. 만성 폐쇄성폐질환은 초기에 자각 증상이 거의 없어, 대부분 중증 상태로 악화된다. 또한, 인종, 성별, 키,몸무게 등 다양한 요인을 포함한 폐 질환 분류 회귀식은 복잡하고, 정확한 판별을 위해서는지속적인 갱신을 필요로 한다. 따라서 폐질환의 초기 진단이 용이하도록 간편한 휴대형 페기능 검사기를 통해 산출 가능한 볼륨-플로우 그래프 이미지 기반 분류 모델이 요구된다. 본 논문에서는 폐질환 조기 진단을 위해 볼륨-플로우 그래프 이미지의 전처리 및 합성곱 신경망 기반 앙상블 딥러닝 모델을 구현하였고, 이를 검증했다. 합성곱 신경망 기반 앙상블 딥러닝 모델은 VGG16, VGG19, ResNet50, 그리고 MobileNet 구조 기반 4개의 모델로 구성되며, 전부 전이학습 및 미세조정하여 사용하였다. 세부적으로는 부족한 수의 학습 데이터를볼륨-플로우 그래프 이미지의 특성을 고려하여 적합한 데이터 증강기법을 적용하였고, 4개의 모델들은 가중치 기반의 간접투표 방식을 사용했다. 최종 앙상블 모델은 단순히 폐질환유무를 판별하는 것이 아닌 정상, 제한성 폐질환, 폐쇄성 폐질환, 그리고 혼합성 폐질환과 같이 총 4개의 클래스로 분류하는 모델임에도 불구하고, 테스트 데이터를 통한 성능은 정확도90.91%, 가중치 평균 정밀도 91.11%, 가중치 평균 재현율 90.91%로 높은 수치를 보였다.

딥러닝 기반 넙치 질병 증상 분류 모델 성능 분석

조경원(Kyung won Cho),백란(Ran Baik) 한국콘텐츠학회 2023 한국콘텐츠학회논문지 Vol.23 No.12

Halibut farming in Korea accounts for more than half of the fishery farming industry. However, 25 to 30 percent of halibut fish deaths are caused by disease per year, which has a very bad effect on the economic feasibility of halibut farming. The accurate diagnosis of halibut disease symptoms in real time is very important for the economic growth of halibut fish farms. In this paper, we propose an independent learning data collection method suitable for a deep learning-based halibut disease symptom classification model, a learning data purification and verification technique that can eliminate labeled learning data set errors, and an equal learning data separation technique, and apply the proposed technique to compare and analyze the halibut disease classification performance for 33 categories of halibut disease symptoms using CNN-based YOLov8 model and Vision Transformer-based Swin model. The YOLOv8 model learned up to 100 Epoch, showing a performance of 0.899 mAP recognition rate, 3 minutes of learning time per 10 Epoch, and 15.4 GB of VRAM usage. The Swin model learned up to 50 Epoch, and the mAP was 0.91, the learning time was 162 minutes per 10 Epoch, and the VRAM usage was 21.3GB. When comparing the performance of the YOLOv8 model and the Swin model, the Swin model showed a good mAP performance recognition rate with less Epoch, but in terms of learning speed, the YOLOv8 model completed the learning with an overwhelmingly short learning time. As shown in the results of this study, if a system that can diagnose halibut disease symptoms in real time using the latest deep learning model is developed, the productivity of the floating halibut style is expected to increase significantly.

Fabrication of a three-dimensional bone marrow niche-like acute myeloid Leukemia disease model by an automated and controlled process using a robotic multicellular bioprinting system

Dana M. Alhattab,Ioannis Isaioglou,Salwa Alshehri,Zainab N. Khan,Hepi H. Susapto,Yanyan Li,Yara Marghani,Arwa A. Alghuneim,Rubén Díaz-Rúa,Sherin Abdelrahman,Shuroug AL-Bihani,Farid Ahmed,Raed I. Felim 한국생체재료학회 2023 생체재료학회지 Vol.27 No.00

Background Acute myeloid leukemia (AML) is a hematological malignancy that remains a therapeutic challenge due to the high incidence of disease relapse. To better understand resistance mechanisms and identify novel therapies, robust preclinical models mimicking the bone marrow (BM) microenvironment are needed. This study aimed to achieve an automated fabrication process of a three-dimensional (3D) AML disease model that recapitulates the 3D spatial structure of the BM microenvironment and applies to drug screening and investigational studies. Methods To build this model, we investigated a unique class of tetramer peptides with an innate ability to selfassemble into stable hydrogel. An automated robotic bioprinting process was established to fabricate a 3D BM (niche-like) multicellular AML disease model comprised of leukemia cells and the BM’s stromal and endothelial cellular fractions. In addition, monoculture and dual-culture models were also fabricated. Leukemia cell compatibility, functionalities (in vitro and in vivo), and drug assessment studies using our model were performed. In addition, RNAseq and gene expression analysis using TaqMan arrays were also performed on 3D cultured stromal cells and primary leukemia cells. Results The selected peptide hydrogel formed a highly porous network of nanofibers with mechanical properties similar to the BM extracellular matrix. The robotic bioprinter and the novel quadruple coaxial nozzle enabled the automated fabrication of a 3D BM niche-like AML disease model with controlled deposition of multiple cell types into the model. This model supported the viability and growth of primary leukemic, endothelial, and stromal cells and recapitulated cell-cell and cell-ECM interactions. In addition, AML cells in our model possessed quiescent characteristics with improved chemoresistance attributes, resembling more the native conditions as indicated by our in vivo results. Moreover, the whole transcriptome data demonstrated the effect of 3D culture on enhancing BM niche cell characteristics. We identified molecular pathways upregulated in AML cells in our 3D model that might contribute to AML drug resistance and disease relapse. Conclusions Our results demonstrate the importance of developing 3D biomimicry models that closely recapitulate the in vivo conditions to gain deeper insights into drug resistance mechanisms and novel therapy development. These models can also improve personalized medicine by testing patient-specific treatments.

Current Status of the Management Program for Animal Model of Disease in an Advanced Country

Yoen Kyung Lee,Seung Eun Jung,Ji Ha Kim,Ji Eun Kim,Hyun Ku Kang,Jung Sik Cho,Jun-Gyo Suh,Dae Youn Hwang 한국실험동물학회 2009 Laboratory Animal Research Vol.25 No.1

Animal models of disease are animals that mimic the pathological condition or the disease entity of specific diseases occurring in humans. These animals are very useful to screen drugs that may be harmful, have a side effect, or would allow a better understanding of the disease mechanism. This study surveyed the current situation of animal care and use program for the Genetic Engineered Mouse (GEM) in the possession of advanced country. Also, the items of these situation involved the GEM number, the national policy and the government support system. The largest number of disease animals in the world is maintained in the USA. Specifically, the National Institute of Health (NIH) in USA supports the several institutions such as Mutant Mouse Regional Resource Centers (MMRRC), Induced Mutant Resource (IMR), Mouse Mutant Gene Resource (MMR), Special Mouse Strains Resource (SMSR) and Neuromice (NMICE) to maintain the animal models of disease. In Japan, RIKEN BioResource Center as one of several resources centers is collecting, supplying and reserving GEM involving the animal models of disease. These works were supported by National BioResource Project (NBRP) of Japanese government. Finally, the some countries in the European Union compile theEuropean Mouse Mutant Archive (EMMA) in order to effectively support and manage these animal models of disease. At present, EMMA possesses 800 kinds of GEM disease models. Therefore, these results suggested that it is very important to establish the national management program for animal model of disease, and this system must be established as soon as possible in our country.

Disease Model의 개발에 관한 기초 연구

양영목,문언수 건국대학교 의과학연구소 1994 건국의과학학술지 Vol.4 No.-

This study was undertaken to develop disease models that investigate effects of environmental factors and estimate genetic parameters for polygenic multifactorial traits, such as hypertension, diabetes, schizophrenia, allergies and cancers. The disease model (1) obtained by the linear model procedure can show significant effects on all environmental factors of the disease, and the disease model (@) can investigate effects of significant environmental factors. The genetic parameter, heritability for the disease can be estimated by regression model (6)) and genetic model (7) using the regression coefficient (b) of offspring on parent(h²=2b).

세계의 구제역 전파,확산역학 모델 개발 현황과 Davis Animal Disease Simulation 모델 국내 적용 연구

윤하정 ( Ha Chung Yoon ),김한 ( Han Kim ),윤순식 ( Soon Seek Yoon ),김연주 ( Youn Ju Kim ),김병한 ( Byoung Han Kim ),( Jack Coleman ),( Tim Carpenter ) 한국예방수의학회(구 한국수의공중보건학회) 2011 예방수의학회지 Vol.35 No.4

Epidemic models on disease spread attempt to simulate disease transmission and associated control processes. This study reviewed published papers on epidemiological models for the management of foot-and-mouth disease in the world. In addition, an individual animal-based, spatially-explicit, stochastic disease transmission model, the Davis Animal Disease Simulation (DADS) model, was described in the frame of an international collaborative research project participating three countries: Republic of Korea, USA, and New Zealand. In this project, the Korean team is aiming at developing the most appropriate parameters for livestock and epidemiology of foot-and-mouth disease outbreaks. On the other hand, the purpose of foreign counterparts is validating their models: DADS (USA) and InterSpread Plus (New Zealand). Classification of farm types and preliminary estimations on the frequency of intra-herd contacts were also presented. This research project is expected to provide precious information to plan a strategy that will facilitate the eradication of foot-and-mouth disease from Korea.

넙치 외형 질병 분류를 위한 딥 매트릭 학습 모델 성능 평가

최한석,강자영,손현승 한국콘텐츠학회 2023 한국콘텐츠학회논문지 Vol.23 No.10

Fish diseases occur year-round in fish farms on the coast of Korea. It is very necessary to diagnose and predict fish diseases in advance for improving the productivity of fish farms. In this paper, we propose a Deep FDC(Fish Diseases Classification) model based on a deep matrix learning model which is a very suitable for classification problem without labeling the learning data of fish disease symptom images. The Deep FDC model shows that learning performance was very high for three diseases classifications(Lymphocystis, Scutica disease, and Vibrioosis). The Recall@K performance evaluation of the Deep FDC model shows that the recall rate of the instance weight loss is 97.14%, the contrast loss is 97.50%, the binomial deviation loss is 96.25%, and the hard mining loss is 96.55%. Moreover, comparing the loss averages for four loss functions in the Deep FDC model, the instance-weighted loss average is 1.83, the contrast loss average is 4.18, the binomial deviation loss average is 1.86, and the lowest loss average is hard mining loss, which is the best learning in the Deep FDC model. The Deep FDC model demonstrates the superiority of using effective loss functions to facilitate disease classification without labeling image segmentation of Paralichthys olivaceus appearance disease symptoms. The expected effect of this paper is to increase the productivity of fish farming as reducing the mass mortality rate caused by diseases by seeing external symptoms and detecting Paralichthys olivaceus diseases early.

System Dynamics Approach to Epidemic Compartment Model: Translating SEIR Model for MERS Transmission in South Korea

정재운 한국디지털정책학회 2018 디지털융복합연구 Vol.16 No.7

Compartment models, a type of mathematical model, have been widely applied to characterize the changes in a dynamic system with sequential events or processes, such as the spread of an epidemic disease. A compartment model comprises compartments, and the relations between compartments are depicted as boxes and arrows. This principle is similar to that of the system dynamics (SD) approach to constructing a simulation model with stocks and flows. In addition, both models are structured using differential equations. With this mutual and translatable principle, this study, in terms of SD, translates a reference SEIR model, which was developed in a recent study to characterize the transmission of the Middle East respiratory syndrome (MERS) in South Korea. Compared to the replicated result of the reference SEIR model (Model 1), the translated SEIR model (Model 2) demonstrates the same simulation result (error=0). The results of this study provide insight into the application of SD relative to constructing an epidemic compartment model using schematization and differential equations. The translated SD artifact can be used as a reference model for other epidemic diseases.

수산화도파민으로 유발된 파킨슨씨병 모델 흰쥐의 행동학 및 면역조직화학적 특성에 관한 연구

장웅규,정천기,오창완,한대희,김현집,조자선,김용식,박찬웅 대한신경외과학회 1998 Journal of Korean neurosurgical society Vol.27 No.2