메뉴보기

KOREAN

ABOUT

PRINCIPAL INVESTIGATOR

PEOPLE

RESEARCH

PUBLICATIONS

NEWS

SEARCH

    HOME

    CURRENT RESEARCH

    • up
    • down

    Improving generalization performance of Natural Language Procesing Models to extract drug safety information from multi-institutional clinical notes

    - To improve the generalizability of a natural language processing model in extracting adverse drug reaction signal from clinical notes.
    - To test the feasibility of using clinical notes generated in diverse clinical settings (e.g., inpatient, outpatient, practitioner specialty, ER visit, operating room) as a means to improve model’s multi-institutional generalizability.
    - To test the applicability of the state-of-the-art generative language models (e.g., GPT4) in improving model robustness.

    An analysis of clinical trial designs for COVID-19 therapeutics development in Korea and suggestions for improvement

    - To analyze COVID-19 clinical trial recommendations from regulatory authorities, academia, and learnings from past successes/failures to propose clinical trial design guidelines for COVID-19 therapeutics.
    - To identify clinical trial design issues by analyzing COVID-19 treatment protocols submitted for funding to Korea Drug Development Fund in Korea.
    - To propose guidelines for efficient clinical trial design to facilitate the timely development of treatments during future public health crises caused by infectious diseases, based on the above analysis.

    Joint research with Microsoft Research Asia (MSRA): Revisiting and reformulating machine learning-based drug repositioning using clinical trials database and eligibility criteria information

    - To highlight the shortcomings of the current approach in using approved drug indication as a prediction target in machine learning-based DR studies
    - To identify a shortcut learning that hinders the generalizability of DR models and propose a more realistic evaluation framework for these models
    - To reframe the DR problem as a natural language processing (NLP) task, using an autoregressive language model like GPT3

    Trajectory of clinical features in Korean COVID-19 patients: an observational study based on real-world data

    - To identify the trajectory of clinical feature in Korean COVID-19 patients using SNUH (Seoul National University Hospital) CDM (Common Data Model) and HIRA (Health Insurance Review & Assessment) CDM
    - To retrospectively analyze changes in diagnoses, medications, test results, and treatments in COVID-19 patients over the course of one year

    Improved prediction of acute coronary syndrome by replacing binary categorical variables with propensity scores in patients with type 2 diabetes mellitus

    - To improve the performance of a risk prediction model in ways that are clinically meaningful and less bias prone
    - To address the problem of false negative and false positive (particularly in drug exposure) in electronic health records (EHRs) when building machine learning models
    - To extend the utility of propensity score beyond covariate balancing to a numerical summary of complex, high dimensional patient health information

    Development of a registry and generation of real-world evidence using electronic medical record-based real-world data for pediatric acute lymphoblastic leukemia

    - To develop a disease registry for pediatric acute lymphoblastic leukemia (ALL) using real-world data (RWD) from the clinical data warehouse (CDW) in multi-center
    - To generate real-world evidence by conducting 3 observational studies for pediatric ALL using electronic medical record (EMR)-based RWD from the common data model (CDM), the CDW, and the registry newly developed in this study
    - To identify opportunities and challenges in the use of real-world data from EMRs to inform regulatory decisions

    CCADD

    Center for Convergence Approaches in Drug Development, Graduate School of Convergence Science and Technology, Seoul National University

    Room C-208, 145 Gwanggyo-ro, Yeongtong-gu, Suwon-si, Gyeonggi-do, 16229, SOUTH KOREA (Gwanggyo)

    Room 406, Building 17, Seoul National University College of Medicine, 103 Daehak-ro, Jongno-gu, Seoul, SOUTH KOREA (Yeon-gun)

    Tel: +82-31-888-9189 (Gwanggyo); +82-2-3668-7381 (Yeon-gun)

    Fax: +82-31-888-9575

    Email: ccadd.snu@gmail.com