곽원건 교수

Programmed death ligand 1 (PD-L1) expression has been associated with clinical outcome of programmed death 1 (PD-1) pathway blockade in non-small cell lung cancer (NSCLC). PD-L1 is highly expressed in approximately 30% of patients without EGFR mutation or ALK rearrangement in previous trials. The objective of this study was to determine the frequency of PD-L1 expression and the clinical outcome according to PD-L1 expression in lung adenocarcinomas harboring EGFR mutation and treated with tyrosin

Background In December 2019, COVID-19 was reported in Wuhan and has since rapidly spread throughout worldwide. In Korea, a large number of COVID-19 patients occurred in Daegu, we, the 3rd hospital in Seoul, began to operate the pre-emptive isolation room for the patients with unexplained fever or pneumonia. In August 2020, the number of COVID-19 patients surged after a large gathering in Seoul. The purpose of this study is to investigate of the patients admitted to the pre-emptive isolation room

<b>Background:</b> High-flow nasal cannula(HFNC) is a recently developed oxygen(O2) supply device, and its use is rapidly increasing in clinical practice. However, studies for factors associated with HFNC outcome are very limited. <b>Aims and Objectives:</b> To evaluate the factors associated with HFNC failure. <b>Methods:</b> For the admitted patients from 1 July 2017 to 30 June 2018 in 5 university affiliated hospitals, we retrospectively reviewed medical records of 1,161 adult patients who we

본 연구의 목적은 다양한 암 환자에서 PD-1 또는 PD-L1 억제제와 관련된 중증 면역 관련 혈액학적 부작용(irHAEs)의 위험을 예측할 수 있는 간단한 머신러닝(ML) 모델을 개발하는 것이다. 우리는 대한민국 대학병원의 전자의무기록 데이터를 기반으로 한 Observational Medical Outcomes Partnership (OMOP) 공통 데이터 모델을 활용하였다. 중증 irHAEs는 Common Terminology Criteria for Adverse Events (CTCAE) 버전 5.0에 따라 3등급에서 5등급으로 정의되었다. 예측 모델은 3차 병원(KHMC)의 주요 데이터 세트를 사용하여 개발되었다. 예측 모델은 특징 중요도 값(FIV)을 기반으로 중요한 특성에 집중하여 최소화되었다. 코호트는 397명의 환자가 포함되었다. 테스트된 ML 알고리즘 중 Random forest가 가장 우수한 예측 성능을 보였으며(AUROC 0.88), FIV의 합이 전체 모델의 5