http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
JoonNyung Heo,Hyungwoo Lee,Il Hyung Lee,Hyo Suk Nam,Young Dae Kim 대한뇌졸중학회 2023 Journal of stroke Vol.25 No.1
Background and Purpose Left atrial or left atrial appendage (LA/LAA) thrombi are frequently observed during cardioembolic evaluation in patients with ischemic stroke. This study aimed to investigate stroke outcomes in patients with LA/LAA thrombus. Methods This retrospective study included patients admitted to a single tertiary center in Korea between January 2012 and December 2020. Patients with nonvalvular atrial fibrillation who underwent transesophageal echocardiography or multi-detector coronary computed tomography were included in the study. Poor outcome was defined as modified Rankin Scale score >3 at 90 days. The inverse probability of treatment weighting analysis was performed. Results Of the 631 patients included in this study, 68 (10.7%) had LA/LAA thrombi. Patients were likely to have a poor outcome when an LA/LAA thrombus was detected (42.6% vs. 17.4%, P<0.001). Inverse probability of treatment weighting analysis yielded a higher probability of poor outcomes in patients with LA/LAA thrombus than in those without LA/LAA thrombus (P<0.001). Patients with LA/LAA thrombus were more likely to have relevant arterial occlusion on angiography (36.3% vs. 22.4%, P=0.047) and a longer hospital stay (8 vs. 7 days, P<0.001) than those without LA/LAA thrombus. However, there was no difference in early neurological deterioration during hospitalization or major adverse cardiovascular events within 3 months between the two groups. Conclusions Patients with ischemic stroke who had an LA/LAA thrombus were at risk of a worse functional outcome after 3 months, which was associated with relevant arterial occlusion and prolonged hospital stay.
기계학습 기법을 이용한 AI 처방의사결정 지원 플랫폼 개발 연구
임태환 ( Tae-hwan Lim ),허준녕 ( Joonnyung Heo ),정혜민 ( Hye-min Jeong ),박삼수 ( Sam-soo Park ),구영환 ( Koo Young-hwan ) 국군의무사령부 2021 대한군진의학학술지 Vol.52 No.1
Objective; Development of Prescription AI decision support system with machine learning Method; An Android based mobile application was developed to provide the user the results from the statistical model we developed. Machine learning model to predict the consumption rate of medical resources was developed with Tensorflow. Result; The solution developed for this study can provide decision support for army medical officers by providing the frequency of prescription for the diagnosis. We can predict the demand for medical supply using the machine learning algorithm developed for this study. The prediction may be helpful in saving costs and meeting supply for demand. Conclusion; The application developed for this study may be not only helpful for increasing accuracy for the management of patients but also be resourceful for the physicians, especially for the first-year army officers or general physicians.