한방진단함수개발에서 진단의 참값으로 사용하는 전문가 진단의 비에 관한 연구

김규곤 한국자료분석학회 2013 Journal of the Korean Data Analysis Society Vol.15 No.6

In developing the diagnosis functions of Oriental Medicine, the true values of diagnosis are made use of medical specialist' diagnosis and the observed values are the data obtained in clinical trial. In this study, we seek for the diagnosis ratio of medical specialist providing the true values of diagnosis functions. Data used in this study are the 3D facial data of clinical trial for 780 cases totally composed of 504 men and 276 women, and five medical specialists are participated in the facial clinical diagnosis. In statistical analysis, we use stepwise discriminant analysis to select analysis variables, and use linear discriminant analysis to obtain the correct rate of diagnosis function. Results of this study in terms of the diagnosis ratio of five medical specialist are as follows. The stable diagnosis functions are derived when diagnoses are consistent with greater than or equal to four specialist' diagnoses in two levels of category, and three specialist' diagnoses in four or five levels of category. 한의학의 진단함수를 개발할 때 진단의 참값으로는 전문가의 진단결과를 사용하고 관측값으로는 임상시험에서 얻은 데이터를 사용하고 있는데, 전문가의 진단이 몇 대 몇으로 일치하는 경우를 진단함수의 참값으로 사용해야 하는지에 대한 전문가의 비를 결정하는 것이 본 연구의 목적이다. 본 연구에서 사용하는 관측값은 임상시험에 참여한 남자 504명, 여자 276명, 합계 780명에 대하여 3차원 얼굴 전용 스캐너를 사용하여 구현된 얼굴형상 데이터이고, 진단의 참값은 5명의 전문가가 얼굴형상을 관찰하여 진단한 결과이다. 3차원 얼굴형상 데이터에 대한 변수선택방법은 단계별판별분석을 이용하고, 형상진단의 진단정확률은 선형판별분석을 이용하고 있다. 본 연구의 결과로부터 전문가 5명의 진단의 비를 요약하면, 형상유형의 범주가 2수준일 경우에는 5:0 또는 4:1로 4명 이상의 진단결과가 일치할 때, 형상유형의 범주가 4수준일 경우에는 5:0:0:0 또는 4:1:0:0 또는 3:2:0:0 또는 3:1:1:0으로 3명 이상의 진단결과가 일치할 때, 형상유형의 범주가 5수준일 경우에는 5:0:0:0:0 또는 4:1:0:0:0 또는 3:2:0:0:0 또는 3:1:1:0:0으로 3명 이상의 진단결과가 일치할 때 안정적인 진단함수를 구할 수 있다.

Factors associated with time to diagnosis from symptom onset in patients with early rheumatoid arthritis

( Soo-kyoung Cho ),( Dam Kim ),( Soyoung Won ),( Jiyoung Lee ),( Chan-bum Choi ),( Jung-yoon Choe ),( Seung-jae Hong ),( Jae-bum Jun ),( Tae-hwan Kim ),( Eunmi Koh ),( Hye-soon Lee ),( Jisoo Lee ),( D 대한내과학회 2019 The Korean Journal of Internal Medicine Vol.34 No.4

Background/Aims: To identify the factors associated with time to diagnosis after symptom onset in patients with early rheumatoid arthritis (RA). Methods: Early RA patients with ≤ 1 year of disease duration in the KORean Observational study Network for Arthritis (KORONA) database were included in this analysis. Patients were further divided into two groups according to the time to diagnosis from symptom onset: the early diagnosis group (time to diagnosis ≤ 1 year) and the late diagnosis group (time to diagnosis > 1 year). Using the multivariable regression model, we identified factors associated with early diagnosis. Results: Among 714 early RA patients, 401 patients (56.2%) and 313 patients (43.8%) were included in the early diagnosis and late diagnosis groups, respectively. The mean disease duration was 0.47 years in the early diagnosis group and 0.45 years in the late diagnosis group. In multivariable model analysis, greater age at onset (odds ratio [OR], 1.03; 95% confidence interval [CI], 1.02 to 1.05), high school education or higher (OR, 1.68; 95% CI, 1.14 to 2.47), higher income (OR, 1.48; 95% CI, 1.05 to 2.08), and initial small joint involvement (OR, 1.42; 95% CI, 1.02 to 1.98) were factors associated with early diagnosis. At diagnosis, disease activity scores using 28 joints on diagnosis (3.81 ± 1.44 vs. 3.82 ± 1.42, p = 0.92) and functional disability (0.65 ± 0.61 vs. 0.57 ± 0.62, p = 0.07) did not different between the two groups. However, hand joint erosion on X-ray (37.8% vs. 25.6%, p < 0.01) was more common in the late diagnosis group than the early diagnosis group. Conclusions: Older onset age, higher educational level and income, and initial small joint involvement were positive factors for early diagnosis of RA.

3개 대학병원의 주 진단 코딩사례 평가

서순원,김광환,부유경,서진숙,서정돈,윤석준,이영성,이무식,정희웅 한국의료QA학회 2002 한국의료질향상학회지 Vol.9 No.1

Background : Coding of principal diagnosis is essential component for producing reliable health statistics. We performed this study to evaluate the current practice of principal diagnoses determination and coding, and to give some basic data to improve coding of principal diagnosis. Method : Nineteen medical record administrators(MRAs) of 3 university hospitals participated in coding principal Dx. From August 1, 2001 to August 31, 2001. From each hospital, 10 medical records of patients with high frequency disease were selected randomly. Each 10 medical records were grouped into three(A,B,C). Then, these 30 medical records were given to each MRAs for coding. At the same time questionnaire was given to each of them. Questions were to prove how they decide and code the principal diagnosis among many current diagnoses ; how they decide and code the principal diagnosis when day see irrelevant diagnosis recorded as the principal diagnosis in medical record, when only tentative diagnoses were recorded without final diagnosis, and when different diagnoses were recorded in different sheets of same record. Agreement of coding among 3 hospitals were compared and survey results were analysed with SAS 6.12. Results : Agreement of coding was found in medical records 5-6 of each 10 medical records. Causes of disagreement were as follows. Difference of clinician’s opinion from each hospital; mixed use of guideline from KCD-3 and guideline from DRG; difference in 4th digit classification according to the absence of pathology report in the medical record; difference of abbreviations among hospitals. 57.9% of decided principal diagnosis after consulting to KCD-3 guideline. When there were difficulties in determining the principal diagnosis, 42.1% of MRAs decided principal diagnosis after discussion with the physician, 26.3% after discussion with fellow MRAs. Conclusion : There were differences in coding among hospitals. To minimize the difference, we suggest the development of disease-specific guidelines for coding in addition to the current general guideline such as KCD-3. To do this, Coding Clinic which can produce guidelines is needed.

Factors associated with time to diagnosis from symptom onset in patients with early rheumatoid arthritis

조수경,김담,원소영,이지영,최찬범,최정윤,홍승재,전재범,김태환,고은미,이혜순,이지수,유대현,윤보영,배상철,성윤경 대한내과학회 2019 The Korean Journal of Internal Medicine Vol.34 No.4

Background/Aims: To identify the factors associated with time to diagnosis after symptom onset in patients with early rheumatoid arthritis (RA). Methods: Early RA patients with ≤ 1 year of disease duration in the KORean Observational study Network for Arthritis (KORONA) database were included in this analysis. Patients were further divided into two groups according to the time to diagnosis from symptom onset: the early diagnosis group (time to diagnosis ≤ 1 year) and the late diagnosis group (time to diagnosis > 1 year). Using the multivariable regression model, we identified factors associated with early diagnosis. Results: Among 714 early RA patients, 401 patients (56.2%) and 313 patients (43.8%) were included in the early diagnosis and late diagnosis groups, respectively. The mean disease duration was 0.47 years in the early diagnosis group and 0.45 years in the late diagnosis group. In multivariable model analysis, greater age at onset (odds ratio [OR], 1.03; 95% confidence interval [CI], 1.02 to 1.05), high school education or higher (OR, 1.68; 95% CI, 1.14 to 2.47), higher income (OR, 1.48; 95% CI, 1.05 to 2.08), and initial small joint involvement (OR, 1.42; 95% CI, 1.02 to 1.98) were factors associated with early diagnosis. At diagnosis, disease activity scores using 28 joints on diagnosis (3.81 ± 1.44 vs. 3.82 ± 1.42, p = 0.92) and functional disability (0.65 ± 0.61 vs. 0.57 ± 0.62, p = 0.07) did not different between the two groups. However, hand joint erosion on X-ray (37.8% vs. 25.6%, p < 0.01) was more common in the late diagnosis group than the early diagnosis group. Conclusions: Older onset age, higher educational level and income, and initial small joint involvement were positive factors for early diagnosis of RA.

Consistency check algorithm for validation and re-diagnosis to improve the accuracy of abnormality diagnosis in nuclear power plants

김근희,김재민,신지현,이승준 한국원자력학회 2022 Nuclear Engineering and Technology Vol.54 No.10

The diagnosis of abnormalities in a nuclear power plant is essential to maintain power plant safety. When an abnormal event occurs, the operator diagnoses the event and selects the appropriate abnormal operating procedures and sub-procedures to implement the necessary measures. To support this, abnormality diagnosis systems using data-driven methods such as artificial neural networks and convolutional neural networks have been developed. However, data-driven models cannot always guarantee an accurate diagnosis because they cannot simulate all possible abnormal events. Therefore, abnormality diagnosis systems should be able to detect their own potential misdiagnosis. This paper proposes a rulebased diagnostic validation algorithm using a previously developed two-stage diagnosis model in abnormal situations. We analyzed the diagnostic results of the sub-procedure stage when the first diagnostic results were inaccurate and derived a rule to filter the inconsistent sub-procedure diagnostic results, which may be inaccurate diagnoses. In a case study, two abnormality diagnosis models were built using gated recurrent units and long short-term memory cells, and consistency checks on the diagnostic results from both models were performed to detect any inconsistencies. Based on this, a re-diagnosis was performed to select the label of the second-best value in the first diagnosis, after which the diagnosis accuracy increased. That is, the model proposed in this study made it possible to detect diagnostic failures by the developed consistency check of the sub-procedure diagnostic results. The consistency check process has the advantage that the operator can review the results and increase the diagnosis success rate by performing additional re-diagnoses. The developed model is expected to have increased applicability as an operator support system in terms of selecting the appropriate AOPs and subprocedures with re-diagnosis, thereby further increasing abnormal event diagnostic accuracy

맥진(脈診)의 사상의학적(四象醫學的) 고찰(考察)

김인태,이수경,이의주,고병희,송일병,Kim, In-Tae,Lee, Soo-Kyung,Lee, Eui-Ju,Koh, Byung-Hui,Song, Il-Byung 사상체질의학회 2003 사상체질의학회지 Vol.15 No.3

The pulse diagnosis is the basic method of oriental medicine diagnosis. But in sasang constitutional medicine, it is said that it's not a major diagnotic method. But we don't have any sasang constitutional study of the pulse diagnosis. So I made this study by research of changing concept of the pulse diagnosis in chinese medical history and Dongyi Suse Bowon (longevity and life preservation in oriental medicine). The conclusion as follows. 1. There were many kinds of the pulse diagnosis in the ancient times, it has developed to simple and effective diagnotic method. Simple and effective is the major point of medical development, the 24 pulse is abridged to Floating Pulse(浮), Deep Pulse(沈), Slow Pulse(遲), Rapid Pulse(數) 2. The latter term of Chosun, the practical study was developed. In the view of the practical study, the pulse diagnosis has a lot of cricical point. Jung Yak-Yong, in his writing Mak Lon(脈論), criticize the pulse diagnosis. 3. In the sasang constitutional medicine, the constitutional diagnosis is very important. The methods of the constitutional diagnosis are three, the way of mind and greed, knowledge and deed, external figure and physical traits. But the pulse diagnosis is one of the way of external figure and physical traits, so we can't diagnose the exact constitution by the pulse diagnosis. 4. Dong-mu conclude that the pulse diagnosis is just the diagnostic way of symptom. But in the clinical situation, the ordinary symptoms are more important than the pulse diagnosis, because it is useful to know the condition of the ingestive food metabolism and the Qi-yack metabolism

사상체질병증(四象體質病症)과 맥진(脈診)의 상관성(相關性)에 대(對)한 문헌적(文獻的) 고찰(考察)

김동준,김정렬,김달래,Kim, Dong-Jun,Kim, Jung-Ryul,Kim, Dal-Rae 대한약침학회 2003 Journal of pharmacopuncture Vol.6 No.3

The purpose of this research was to investigate the correlation Among Sasang Constitutional Disease and Examination of the pulse. I have gone over literatures of mainly ${\ulcorner}$Dongyi Soose Bowon${\lrcorner}$ and the others Oriental Medical book was studied about the Pulse Diagnosis. And then I came to get some conclusion as follows. 1. Soeumin(소음인) the initial-stage symptoms of wulkwang disease(울광증) ; when the Superficial Pulse and the Superficial+ Moderate Pulse is made a diagnosis, Ceongunggyegitang(천궁계지탕) and Gunggyuhyangsosan(궁귀향소산) can be used. 2. Soeumin(소음인) the initial-stage blood disease symptoms of wulkwang disease(울광증) ; when the Minute+deep Pulse is made a diagnosis, Palmulgnnjatang(팔물군자탕) and Guakhyanggeonggisan(곽향정기산) can be used. 3. Soeumin(소음인) the initial-stage symptoms of mangyang disease(망양증) ; when the Yang region Superficial Pulse and the Yin region Weak Pulse is made a diagnosis, Hwanggigyegitang(황기계지탕), Bojungikgitang(보증익기탕) and Sengyangikgitang(승양익기탕) can be used. 4. Soeumin(소음인) the symptoms of taeum disease(태음증) ; when the Minute Pulse and Deep+Thin Pulse is made a diagnosis, Sasang Prescription can be used. 5. Soeumin(소음인) the symptoms of soeum disease(소음증) ; when the Minute+Thin Pulse, Deep Pulse and Thin+Deep+Rapid Pulse is made a diagnosis, Sasang Prescription can be used. 6. Soyangin(소양인) Wind of soyang disease(소양상풍증) ; when the Superficial+Tight Pulse is made a diagnosis, Hungbangpaedogsan(형방패독산) can be used. And when the Deep+Full with strong power Pulse is made a diagnosis, Hyungbangdojeoksan(형방도적산) can be used. 7. Soyangin(소양인) the symptoms of mangyeum disease(망음증) ; when the Superficial+Large+Rapid Pulse and Flood+Large Pulse is made a diagnosis, Hungbangsabaeksan(형방사백산) can be used. And when the Wiry+Thin Pulse is made a diagnosis, Hungbanggiwhangtang(형방지황탕) can be used. 8. Soyangin(소양인) the chest-phrenic fever syndrome(흉격열증) ; when the Superficial Pulse, Flood+Full+Rapid Pulse and Flood+Large Pulse is made a diagnosis, Sasang Prescription can be used. 9. Soyangin(소양인) the after fever syndrome(음허오열증) ; when the Empty+Soft+Rapid Pulse is made a diagnosis, Sasang Prescription can be used. 10. Taeumin(태음인) the upper neck exterior disease caused by Cold(배추표병) ; when the Superficial and Superficial+Tight Pulse is made a diagnosis, Mawhangbalpoytang(마황발표탕) can be used, And when the Superficial and Superficial+Tight with strong power on left hand Pulse is made a diagnosis, Ungdamsan(웅담산) and Handayulsotang(한다열소탕) can be used. 11. Taeumin(태음인) the Coldness syndrome in esophagus(위완한증) ; when the Superficial+Tight Pulse with weak power on left hand Pulse is made a diagnosis, Taeumjowetang(태음조위탕) can be used. 12. Taeumin(태음인) the Dryness-Heat syndrome(조열증) ; when the Flood+Large Pulse, Long Pulse and Long+Large Pulse is made a diagnosis, Galgeunhaegitang(갈근해기탕) can be used. And when the Tight+Full+Rapid Pulse with deep region is made a diagnosis, Yuldahansotang(열다한소탕) can be used. And when the Superficial+Slippery Pulse is made a diagnosis, Chungsimyunjatang(청심연자탕) can be used. 13. Taeumin(태음인) the symptoms of Yin-blood Exhaustion(음혈모갈증) ; when the Superficial with weak power Pulse is made a diagnosis, Nokyongdaebotang(녹용대보탕) can be used. And when the Deep with weak power Pulse is made a diagnosis, Gongjinheukwondan(공진흑원단) can be used. 14. Taeyangin(태양인) a slight Lumbar vertebrae disease(외감경증) ; when the Superficial+Hollow Pulse is made a diagnosis, Gunshitang(건시탕) can be used. 15. Taeyangin(태양인) the Generalized and Fatigue syndrome(해역증) ; when the Moderate+Choppy Pulse with left

무균성 뇌막염에서 증상발현부터 진단까지 걸린 시간에 따른 시기별 유병기간의 검토

김탁수,허지연,박영희,정민구,김성원,Kim, Tag Soo,Hur, Ji Yeon,Park, Young Hee,Jung, Min Goo,Kim, Sung Won 대한소아감염학회 1996 Pediatric Infection and Vaccine Vol.3 No.2

Purpose : Aseptic meningitis is relatively frequent in children and caused mostly by enterovirus. The aim of the present study was to determine the effect of early diagnosis (spinal tapping) on symptom duration of childhood aseptic meningitis. Methods : One hundred fifty-three children who were hospitalized due to aseptic menigitis in the Department of Pediatrics St. Benedict Hospital from July 1996 through October 1996 were included in this study. Patients were divided to two groups according to the duration from first symptom onset to diagnosis. Early diagnosis group is diagnosed within 3 days from first symptom onset. Later diagnosis group is diagnosed after 4 days from first symptom onset. Results : 1) The average age of these patients was 4.3 years old in early diagnosis group and 4.1 years old in later diagnosis group. The sex ratio(male: female) was 2.04:1 in early diagnosis group and 2.5:1 in later diagnosis group. 2) The mean duration of diagnosis of this study was 2.04 day in early diagnosis group and 5.12 day in later diagnosis group. 3) The percentage of symptom and sign of the early diagnosis group were fever(100%), headache(88.4%), vomiting(86.9%), abdominal pain(39%), neck stiffness(36.2%), skin rash(18.8%), diarrhea(16.9%) and that of later diagosis group were fever(100%), headache(83.3), vomiting(80.9%), abdominal pain(47.6%), neck stiffness(41.6%), skin rash(29.7%), diarrhea(16.6%). 4) Initial CSF findings revealed leukocyte $146.8{\pm}386.3/mm^3$ with PMNL 38%, protein 32.47mg/dl, sugar 66.23mg/dl in early diagnosis group and leukocyte $458.1{\pm}663.2/mm^3$, protein 31.22mg/dl, sugar 64.21 mg/dl in later diagnosis group. 5) There was no statistically significant differance in the peripheral blood findings between early diagnosis group and later diagnosis group. 6) The duration of disappearance of symptom after spinal tap were 2.3 days in early diagnosis group and 2.24 days in later diagnosis group. Total symptom duration was 4.34 days in early diagnosis group and 7.36 days in later diagnosis group. Conclusions : Our results demonstrate that early diagnosis(early spinal tap) shortened duration of clinical symptoms.

정부 조직진단의 추진전략과 관리방안

김윤권(Yunkwon Kim) 한국공공관리학회 2007 한국공공관리학보 Vol.21 No.4

본 연구는 공공부문의 조직진단이 적실성과 타당성을 결여하고 있다는 문제의식에서 조직진단과 관련된 주요 이론적 논점을 부각시키고, 바람직한 정부조직진단 결과를 도출하기 위한 조직진단 추진전략과 관리방안을 제시한다. 바람직한 조직진단 추진전략으로는 진단목적 공유, 진단의 정확한 인과성 확보, 상시적 진단체제 구축, 기관별 맞춤형 진단 구축, 협력형 진단체제 구축을 제시한다. 또한 타당성과 적실성을 지속적으로 확보하기 위한 조직진단 관리방안으로는 조직 리더의 적극적인 역할, 조직진단실무자의 전문적 역할, 진단모듈별 상호 연계된 진단관리, 조직진단 기법 및 지표 고도화, 조직진단 결과의 평가 및 활용을 제시한다. 정부 조직진단에 대한 이론적 검토, 추진전략, 관리방안은 최근 부각되고 있는 조직진단의 정체성 및 내용을 이해시키고 조직진단의 타당성, 적실성, 및 수용성을 제고시켜 줄 것이다. 일반성과 특수성을 고려한 정부조직진단을 정교화시켜 향후 행정개혁이나 조직개편에 유용한 대안을 제시하고 궁극적으로는 행정서비스를 높이고 국가경쟁력을 높이는 방향으로 조직진단이 이루어져야 한다는 당위성을 제시한다. Under the research question that public organizational diagnosis needs relevance and validity, this paper focuses on theoretical issues on organizational diagnosis. And it suggests the strategies and management for organizational diagnosis in order to induce the desirable results of government organizational diagnosis. The desirable strategies for organizational diagnosis will be suggested as follows: sharing diagnosis goals, acquiring cause and effect of diagnosis, building persistent diagnosis, setting custom-made diagnosis for institutions, and building cooperative diagnosis system. The management method of diagnosis for persistent validity and relevance in government organizational will be recommended as follows: organizational leader's active role, diagnosis practitioner's professional role, organic diagnosis relationship among diagnosis modules, enhancing the manuel and index for organizational diagnosis, the evaluation and usage of the diagnosis results. The review of theoretical diagnosis, the diagnosis strategies and the diagnosis management will clear the identity and contents of organizational diagnosis. And it will improve the validity, relevance and acceptance for organizational diagnosis. This paper will pave the way for usable alternatives for administrative reform or reorganization by elaborating the government organizational diagnosis considering generalization and specificity. This paper will imply the legitimacy of organizational diagnosis by enhancing public administrative service and state competency.

웹취약점 자동진단 개선방안

김태섭,조인준 한국콘텐츠학회 2022 한국콘텐츠학회논문지 Vol.22 No.2

Due to the development of smartphone technology, as of 2020, 91.9% of people use the Internet[1] to frequently acquire information through websites and mobile apps. As the number of homepages in charge of providing information is increasing every year, the number of applications for web vulnerability diagnosis, which diagnoses the safety of homepages, is also increasing. In the existing web vulnerability check, the number of diagnostic personnel should increase in proportion to the number of homepages that need diagnosis because the diagnosticians manually test the homepages for vulnerabilities. In reality, however, there is a limit to securing a web vulnerability diagnosis manpower, and if the number of diagnosis manpower is increased, a lot of costs are incurred. To solve these problems, an automatic diagnosis tool is used to replace a part of the manual diagnosis. This paper explores a new method to expand the current automatic diagnosis range. In other words, automatic diagnosis possible items were derived by analyzing the impact of web vulnerability diagnosis items. Furthermore, automatic diagnosis identified possible items through comparative analysis of diagnosis results by performing manual and automatic diagnosis on the website in operation. In addition, it is possible to replace manual diagnosis for possible items, but not all vulnerability items, through the improvement of automatic diagnosis tools. This paper will explore some suggestions that can help improve plans to support and implement automatic diagnosis. Through this, it will be possible to contribute to the creation of a safe website operating environment by focusing on the parts that require precise diagnosis. 스마트폰 기술의 발전으로 인하여 2020년 기준 전 국민의 91.9%가 인터넷 이용[1]하여 수시로 홈페이지와 모바일 앱을 통해 정보를 습득하고 있다. 정보제공을 담당하는 홈페이지의 수가 점점 늘어남에 따라 홈페이지의 안전성을 진단하는 웹취약점 진단 신청수도 매년마다 증가하고 있는 상황이다. 기존 웹취약점 점검은 진단원이 수작업으로 홈페이지를 모의 해킹하여 취약점을 진단했기 때문에 진단대상 홈페이지 수에 비례하여 진단인력이 늘어나야한다. 하지만 현실적으로 웹취약점 진단인력 확보에 한계가 있고, 진단인력을 늘렸을 경우 많은 비용이 발생한다. 이러한 문제점 해결을 위해 자동진단 도구를 사용하여 수동진단의 일부를 대체하고 있다. 본 논문에서는 현재의 자동진단 범위를 확대하기 위한 방안을 새롭게 제안하였다. 즉, 웹취약점 진단항목의 영향도를 분석하여 자동진단 가능 항목을 도출하고, 실제 운영 중인 홈페이지에 수동 및 자동진단을 수행하여 진단결과에 대한 비교 분석을 통해 자동진단 가능항목을 파악하였다. 또한 자동진단 개선방안을 제시하여 자동진단 도구 개선을 통해 모든 취약점 항목은 아니지만 가능한 항목에 대해서 수동진단을 대체할 수 있다. 이를 통해서 진단 및 정밀진단이 필요한 부분에 집중하여 안전한 홈페이지 운영환경 조성에 기여할 수 있을 것이다.