• 대한의용생체공학회:의공학회지
• /
• 제5권2호
• /
• pp.121-126
• /
• 1984

A new approach to texture classification for quantitative ultrasound liver diagnosis using run difference matrix was developed. The run difference matrix comprised the gray level difference along with a distances. From this run difference matrix, we defined several vectors and parameters such as DOD, DGD, DAD vector, SHP, SMO, SMG, LDE, LDEL etc.Each parameter values calculated in fatty, cirrhotic, normal and chronic hepatitic liver images were plotted in a plane and we found that RDM method was more sensitive to small structural changes than the conventional run length method and showed improved classification ability between the diseases.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2002년도 하계학술대회 논문집 D
• /
• pp.2815-2817
• /
• 2002

Classification method of cancers using cDNA microarrays data was developed using genetic algorithms and neural networks. For gene selection, 2308 genes were ranked using genetic algorithms, and selected by frequency number of selection from 1000 of genetic iterative runs. To calculate fitness values, artificial neural networks are used as classifier. The small, round blue cell tumors (SRBCTs) which is difficult to distinguish via pathological single test was used as test diseases for classification, and the test results showed the 96% of exact classification capability for 25 test samples.

• 대한의용생체공학회:의공학회지
• /
• 제43권2호
• /
• pp.109-115
• /
• 2022

Purpose: Medical records classification using vectorization techniques plays an important role in natural language processing. The purpose of this study was to investigate proper vectorization techniques for electronic medical records classification. Material and methods: 403 electronic medical documents were extracted retrospectively and classified using the cosine similarity calculated by Scikit-learn (Python module for machine learning) in Jupyter Notebook. Vectors for medical documents were produced by three different vectorization techniques (TF-IDF, latent sematic analysis and Word2Vec) and the classification precisions for three vectorization techniques were evaluated. The Kruskal-Wallis test was used to determine if there was a significant difference among three vectorization techniques. Results: 403 medical documents were relevant to 41 different diseases and the average number of documents per diagnosis was 9.83 (standard deviation=3.46). The classification precisions for three vectorization techniques were 0.78 (TF-IDF), 0.87 (LSA) and 0.79 (Word2Vec). There was a statistically significant difference among three vectorization techniques. Conclusions: The results suggest that removing irrelevant information (LSA) is more efficient vectorization technique than modifying weights of vectorization models (TF-IDF, Word2Vec) for medical documents classification.

• 보건행정학회지
• /
• 제11권1호
• /
• pp.1-18
• /
• 2001

The purpose of the present study was to assess the agreement of survival probability estimated by International Classification of Diseases l0th Edition(ICD-10) based International Classification of Diseases based Injury Severity Score(ICISS) with professional panel's judgment on preventable death. ICISS has a promise as an alternative to Trauma and Injury Severity Score(TRISS) which have served as a standard measure of trauma severity, but requires more validation studies. Furthermore as original version of ICISS was based ICD-9CM, it is necessary to test its performance employing ICD-10 which has been used in Korea and is expected to replace ICD-9 in many countries sooner or later. Methods : For 1997 and 1998 131 trauma deaths and 1,785 blunt trauma inpatients from 6 emergency medical centers were randomly sampled and reviewed. Trauma deaths were reviewed by professional panels with hospital records and survival probability of trauma inpatients was assessed using ICD-10 based ICISS. For trauma mortality degree of agreement between ICISS survival probability with judgment of professional panel on preventable death was assessed and correlation between W-score and preventable death rate by each emergency medical center was assessed. Results : Overall agreement rate of ICISS survival probability with preventable death judged by professional panel was 66.4%(kappa statistic 0.36). Spearman's correlation coefficient between W-score and preventable death rate by each emergency medical center was -0.77(p=0.07) and Pearson's correlation coefficient between them was -0.90(p=0.01). Conclusions : The agreement rate of ICD-10 based ICISS survival probability with of professional panel's judgment on preventable death was similar to TRISS. The W-scores of emergency medical centers derived from ICD-10 based ICISS were highly correlated with preventable death rates of them with marginal statistical significance.

• Journal of Preventive Medicine and Public Health
• /
• 제51권1호
• /
• pp.15-22
• /
• 2018

Objectives: The use of administrative data is an affordable alternative to conducting a difficult large-scale medical-record review to estimate the scale of adverse events. We identified adverse events from 2002 to 2013 on the national level in Korea, using International Classification of Diseases, tenth revision (ICD-10) Y codes. Methods: We used data from the National Health Insurance Service-National Sample Cohort (NHIS-NSC). We relied on medical treatment databases to extract information on ICD-10 Y codes from each participant in the NHIS-NSC. We classified adverse events in the ICD-10 Y codes into 6 types: those related to drugs, transfusions, and fluids; those related to vaccines and immunoglobulin; those related to surgery and procedures; those related to infections; those related to devices; and others. Results: Over 12 years, a total of 20 817 adverse events were identified using ICD-10 Y codes, and the estimated total adverse event rate was 0.20%. Between 2002 and 2013, the total number of such events increased by 131.3%, from 1366 in 2002 to 3159 in 2013. The total rate increased by 103.9%, from 0.17% in 2002 to 0.35% in 2013. Events related to drugs, transfusions, and fluids were the most common (19 446, 93.4%), followed by those related to surgery and procedures (1209, 5.8%) and those related to vaccines and immunoglobulin (72, 0.3%). Conclusions: Based on a comparison with the results of other studies, the total adverse event rate in this study was significantly underestimated. Improving coding practices for ICD-10 Y codes is necessary to precisely monitor the scale of adverse events in Korea.

• Tuberculosis and Respiratory Diseases
• /
• 제87권3호
• /
• pp.221-233
• /
• 2024

The 2023 Global Initiative for Chronic Obstructive Lung Disease (GOLD) revised the definition of chronic obstructive pulmonary disease (COPD) to broadly include a variety of etiologies. A new taxonomy, composed of etiotypes, aims to highlight the heterogeneity in causes and pathogenesis of COPD, allowing more personalized management strategies and emphasizing the need for targeted research to understand and manage COPD better. However, controversy arises with including some diseases under the umbrella term of COPD, as their clinical presentations and treatments differ from classical COPD, which is smoking-related. COPD due to infection (COPD-I) and COPD due to environmental exposure (COPD-P) are classifications within the new taxonomy. Some disease entities in these categories show distinct clinical features and may not benefit from conventional COPD treatments, raising questions about their classification as COPD subtypes. There is also controversy regarding whether bronchiectasis with airflow limitations should be classified as an etiotype of COPD. This article discusses controversial issues associated with the proposed etiotypes for COPD in terms of COPD-I, COPD-P, and bronchiectasis. While the updated COPD definition by GOLD 2023 is a major step towards recognizing the disease's complexity, it also raises questions about the classification of related respiratory conditions. This highlights the need for further research to improve our understanding and approach to COPD management.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2004년도 학술대회 논문집 정보 및 제어부문
• /
• pp.325-327
• /
• 2004

ECG limb lead II signal widely used to diagnosis heart diseases and it is essential to detect ECG events (onsets, offsets and peaks of the QRS complex P wave and T wave) and extract them from ECG signal for heart diseases diagnoses. However, it is very difficult to develop standardized feature extraction formulas since ECG signals are varying on patients and disease types. In this paper, simple feature extraction method from normal and abnormal types of ECG signals is proposed. As a signal features, heart rate, PR interval, QRS interval, QT interval, interval between S wave and baseline, and T wave types are extracted. To show the validity of proposed method, Right Bundle Branch Block (RBBB), Left Bundle Branch Block (LBBB), Sinus Bradycardia, and Sinus Tachycardia data from MIT-BIH arrhythmia database are used for feature extraction and the extraction results showed higher extraction capability compare to conventional formula based extraction method.

• Tuberculosis and Respiratory Diseases
• /
• 제83권2호
• /
• pp.122-131
• /
• 2020

Hypersensitivity Pneumonitis (HP) one of the most common interstitial lung diseases (ILDs) is characterized by exposure to an inhaled inciting antigen that leads to a host immunologic reaction determining interstitial inflammation and architectural distortion. The underlying pathogenetic mechanisms are unclear. The absence of international shared diagnostic guidelines and the lack of a "gold-standard" test for HP combined with the presence of several clinical and radiologic overlapping features makes it particularly challenging to differentiate HP from other ILDs, also in expert contests. Radiology is playing a more crucial role in this process; recently the headcheese sign was recognized as a more specific for chronic-HP than the extensive mosaic attenuation. Several classification proposals and diagnostic models have been advanced by different groups, with no prospective validation. Therapeutic options for HP have been limited to antigen avoidance and immunosuppressant drugs over the last decades. Several questions about this condition remain unanswered and there is a need for more studies.

• 대한예방한의학회지
• /
• 제17권2호
• /
• pp.47-68
• /
• 2013

Objectives : This research aimed to develop a new etiologic classification for traditional Korean Medicine in order to respond to the social and environmental change. Methods : We reviewed the existing theories on etiological classification for East Asian Medicine thoroughly and discussed the problems and limitations. Based on the experts' consensus, we abstracted disease factors and etiologic items. Results : The disease factors are classified into three parts: the human body, the environment, and the interaction between the human body and the environment. We defined them as the inner factor, the external factor, and the interaction between the inner and the external factors. The inner factor is free from the influence of the environment, and it causes diseases solely from the components of the human body. It is divided into genetic factors. The external factor is defined as a case when a disease occurs due to a factor outside the human body and includes external injuries, environmental pollution, and natural disasters. The interaction between the inner and the external factors is a disease factor that causes diseases by the interaction of the human body and the environment and includes emotions, habits, and social environment. As a result of the analysis, it was possible to see the meanings at a single glance as the scattered and fractional meanings were integrated with focus on medicinal herbs, but the increasing number of analyzed medicinal herbs tended to more and more complicate their relationships, thus, requiring additional work like filtering. Conclusions : The new etiologic classification of Korean Medicine fully reflects the perspectives on life in Korean Medicine while embracing the changes in modem society. Also, by avoiding the usage of ambivalent terms and wrong classification methods, the new classification system constructs intuitive and concise etiology and improves usability in clinical medicine.

• 동의생리병리학회지
• /
• 제32권4호
• /
• pp.217-225
• /
• 2018

This study was to investigate the relationship between ultrasonographic examination and pattern identification classification on cinical studies in chinese medicine. We searched clinical studies related correlation between ultrasonographic examination and pattern identification classification in chinese medicine, that published from 2013 to 2016 in China National Knowledge Infrastructure (CNKI) databases by keywords, 'ultrasound(超?)', 'chinese medicine(中?)', 'syndrome differentiation (辨?)'. Seventeen studies were found. There were 7 studies of gynecological diseases including polycystic ovary syndrome and uterine myoma, 5 studies of fatty liver, 3 studies of arthritis, and 1 studie of thyroid nodule and lymphadenopathy respectively. As a result, ii is thought that there was a certain degree of correlation between the change of the ultrasonographic image and the pathological types according to traditional chinese medicine (TCM) syndrome differentiation and ultrasonographic examination could be used as secondary means for the TCM syndrome differentiation classification. In conclusion, by using ultrasonograph device in a medicinal way of TCM and traditional korean medicine (TKM), it is thought that more detailed and accurate diagnosis and treatment are possible and the evidence for reasonableness of syndrome differentiation in TCM and TKM its validity can be secured.