• 한국항행학회논문지
• /
• 제25권1호
• /
• pp.96-101
• /
• 2021

본 논문은 빅데이터를 이용하여 심방세동 환자의 뇌졸중 발병을 예측하는 기계 학습 모델을 제시한다. 학습 데이터로는 국민 건강 보험공단에서 제공하는 대한민국 전수에 해당하는 심방세동 환자의 정보를 수집하였다. 수집된 정보는 인구사회학, 과거 병력, 건강검진을 포함한 68개 독립변수로 구성된다. 본 연구의 목표는 기존 심방세동 환자의 뇌졸중 위험도 예측에 사용되던 통계적 모델 (CHADS2, CHA2DS2-VASc)의 성능을 검증하고 기계 학습 모델을 적용하여 기존 모델보다 높은 정확도를 가지는 모델을 제시하는 것이다. 제안하는 모델의 정확도, AUROC (area under the receiver operating characteristic)를 검증한 결과 제안하는 기계 학습 기반의 모형이 심방세동 환자의 뇌졸중 위험도를 사용한 모델이 기존의 통계적 모델보다 높은 정확도, 민감도, 특이도를 가지는 것을 확인할 수 있었다.

• 한국컴퓨터정보학회논문지
• /
• 제27권12호
• /
• pp.69-76
• /
• 2022

본 논문에서는 MIMIC-IV(Medical Information Mart for Intensive Care) v2.0 데이터를 이용한 시계열 데이터의 정제 및 가공 방법을 제안한다. 더불어 해당 가공법을 기반으로 정제한 데이터셋을 활용하여 구축한 기계학습 기반의 욕창 조기 경보 시스템을 통해 해당 가공 방법의 유의성을 검증하였다. 구현된 욕창 조기 경보 시스템은 병변이 발생하기 전 12, 24시간에 미리 의료진에게 경보를 주는 시스템이다. 전자의무기록(Electronic Medical Record; EMR) 시스템과 연동하여 실시간으로 환자의 욕창 발생 위험도를 의료진에게 알려 중환자 의사결정을 지원하고, 나아가 효율적인 의료 자원 배분을 가능하게 한다. 여러 기계학습 모델 중 GRU 모델을 사용하였을 때, AUROC 평가지표를 기준으로 발생 전 12시간이 0.831, 24시간이 0.822로 가장 좋은 성능을 보였다.

• Korean Journal of Radiology
• /
• 제25권1호
• /
• pp.74-85
• /
• 2024

Objective: Idiopathic intracranial hypertension (IIH) is a condition of unknown etiology associated with venous sinus stenosis. This study aimed to develop a magnetic resonance venography (MRV)-based radiomics model for predicting a high trans-stenotic pressure gradient (TPG) in IIH patients diagnosed with venous sinus stenosis. Materials and Methods: This retrospective study included 105 IIH patients (median age [interquartile range], 35 years [27-42 years]; female:male, 82:23) who underwent MRV and catheter venography complemented by venous manometry. Contrast enhanced-MRV was conducted under 1.5 Tesla system, and the images were reconstructed using a standard algorithm. Shape features were derived from MRV images via the PyRadiomics package and selected by utilizing the least absolute shrinkage and selection operator (LASSO) method. A radiomics score for predicting high TPG (≥ 8 mmHg) in IIH patients was formulated using multivariable logistic regression; its discrimination performance was assessed using the area under the receiver operating characteristic curve (AUROC). A nomogram was constructed by incorporating the radiomics scores and clinical features. Results: Data from 105 patients were randomly divided into two distinct datasets for model training (n = 73; 50 and 23 with and without high TPG, respectively) and testing (n = 32; 22 and 10 with and without high TPG, respectively). Three informative shape features were identified in the training datasets: least axis length, sphericity, and maximum three-dimensional diameter. The radiomics score for predicting high TPG in IIH patients demonstrated an AUROC of 0.906 (95% confidence interval, 0.836-0.976) in the training dataset and 0.877 (95% confidence interval, 0.755-0.999) in the test dataset. The nomogram showed good calibration. Conclusion: Our study presents the feasibility of a novel model for predicting high TPG in IIH patients using radiomics analysis of noninvasive MRV-based shape features. This information may aid clinicians in identifying patients who may benefit from stenting.

• Korean Journal of Radiology
• /
• 제23권9호
• /
• pp.921-930
• /
• 2022

Objective: To identify epidermal growth factor receptor (EGFR) mutations in lung adenocarcinoma based on ¹⁸F-fluorodeoxyglucose (FDG) PET/CT radiomics and clinical features and to distinguish EGFR exon 19 deletion (19 del) and exon 21 L858R missense (21 L858R) mutations using FDG PET/CT radiomics. Materials and Methods: We retrospectively analyzed 179 patients with lung adenocarcinoma. They were randomly assigned to training (n = 125) and testing (n = 54) cohorts in a 7:3 ratio. A total of 2632 radiomics features were extracted from the tumor region of interest from the PET (1316) and CT (1316) images. Six PET/CT radiomics features that remained after the feature selection step were used to calculate the radiomics model score (rad-score). Subsequently, a combined clinical and radiomics model was constructed based on sex, smoking history, tumor diameter, and rad-score. The performance of the combined model in identifying EGFR mutations was assessed using a receiver operating characteristic (ROC) curve. Furthermore, in a subsample of 99 patients, a PET/CT radiomics model for distinguishing 19 del and 21 L858R EGFR mutational subtypes was established, and its performance was evaluated. Results: The area under the ROC curve (AUROC) and accuracy of the combined clinical and PET/CT radiomics models were 0.882 and 81.6%, respectively, in the training cohort and 0.837 and 74.1%, respectively, in the testing cohort. The AUROC and accuracy of the radiomics model for distinguishing between 19 del and 21 L858R EGFR mutational subtypes were 0.708 and 66.7%, respectively, in the training cohort and 0.652 and 56.7%, respectively, in the testing cohort. Conclusion: The combined clinical and PET/CT radiomics model could identify the EGFR mutational status in lung adenocarcinoma with moderate accuracy. However, distinguishing between EGFR 19 del and 21 L858R mutational subtypes was more challenging using PET/CT radiomics.

• Clinical Endoscopy
• /
• 제55권1호
• /
• pp.122-127
• /
• 2022

Background/Aims: Endoscopic ultrasonography (EUS) is warranted when cross-sectional imaging demonstrates common bile duct (CBD) dilatation without identifiable causes. This study aimed to assess the diagnostic performance of EUS in CBD dilatation of unknown etiology. Methods: Retrospective review of patients with dilated CBD without definite causes undergoing EUS between 2012 and 2017. Results: A total of 131 patients were recruited. The mean age was 63.2±14.1 years. The most common manifestation was abnormal liver chemistry (85.5%). The mean CBD diameter was 12.2±4.1 mm. The area under the receiver operating characteristic curve (AUROC) of EUS-identified pathologies, including malignancy, choledocholithiasis, and benign biliary stricture (BBS), was 0.98 (95% confidence interval [CI], 0.95-1.00). The AUROC of EUS for detecting malignancy, choledocholithiasis, and BBS was 0.91 (95% CI, 0.85-0.97), 1.00 (95% CI, 1.00-1.00), and 0.93 (95% CI, 0.87-0.99), respectively. Male sex, alanine aminotransferase ≥3× the upper limit of normal (ULN), alkaline phosphatase ≥3× the ULN, and intrahepatic duct dilatation were predictors for pathological obstruction, with odds ratios of 5.46 (95%CI, 1.74-17.1), 5.02 (95% CI, 1.48-17.0), 4.63 (95% CI, 1.1-19.6), and 4.03 (95% CI, 1.37-11.8), respectively. Conclusions: EUS provides excellent diagnostic value in identifying the etiology of CBD dilatation detected by cross-sectional imaging.

• The Korean journal of internal medicine
• /
• 제39권2호
• /
• pp.248-260
• /
• 2024

Background/Aims: The model for end-stage liver disease (MELD) serves as an indicator for short-term mortality among patients diagnosed with liver cirrhosis (LC) and is used to prioritize patients for liver transplantation. In 2021, the updated version of MELD, MELD-3.0, was introduced to improve the accuracy of the mortality prediction of MELD. Therefore, this study aimed to compare the efficacy of MELD 3.0 and MELD-Na in predicting mortality among Korean patients with LC. Methods: A retrospective review was conducted using the medical records of patients diagnosed with LC who were admitted to Konkuk University Hospital From 2011 to 2021. The study calculated the predictive values of MELD-Na and MELD-3.0 for 3- and 6-months mortality using the area under the receiver operating curve (AUROC) and compared the results using the DeLong test. Results: Of the 3,034 patients enrolled in the study, 339 (11.2%) died within 3 months and 421 (14.4%) died within 6 months. The AUROCs values for predicting 3 months mortality were 0.846 for MELD-Na and 0.851 for MELD-3.0. The corresponding AUROC values for predicting 6 months mortality were 0.843 for MELD-Na and 0.848 for MELD-3.0. MELD-3.0 exhibited better discrimination ability than MELD-Na for both 3 (p = 0.03) and 6 months mortality (p = 0.01). Conclusions: Our study found a significant difference between the performance of MELD-3.0 and MELD-Na in Korean patients with LC.

• 한국산림과학회지
• /
• 제112권4호
• /
• pp.426-441
• /
• 2023

본 연구에서는 수확벌채에 따른 수목의 뿌리 점착력의 변화와 토양의 포화를 가정한 지표유출의 세 가지 흐름 기법(SFD; Single flow direction, MFD; Multiple flow direction, IFD; Infinite flow direction)을 무한사면 안전율 공식에 적용하여 산사태 발생 예측 모델링의 정확성을 분석하였다. 이를 위해 2020년 8월 집중호우의 영향으로 자연사면과 벌채사면에서 다수의 산사태가 발생한 제천지역을 연구지역으로 선정하였다. 위성영상과 25cm급 항공사진을 이용한 산사태 인벤토리 맵핑 결과, 연구지역 내에서 총 830개소의 산사태 발생원이 확인되었다. 산사태 모델링 결과, 벌채에 따른 뿌리 점착력의 변화를 고려한 경우(MFD: 0.81, IFD: 0.80, SFD: 0.80)가 벌채의 영향을 고려하지 않은 경우(MFD: 0.79, IFD: 0.79, SFD: 0.78)에 비하여 AUROC(Area Under the Receiver Operating Characteristics) 분석에서 정확성이 1.3~2.6% 향상되는 것으로 나타났다. 또한, MFD 알고리즘을 이용한 경우는 다른 알고리즘과 비교하여 AUROC 분석에서 정확성이 최대 1.3% 향상되었다. 이러한 결과는 식생조건의 변화를 고려한 뿌리 점착력의 차등 적용과 지표유출수 흐름기법의 선정이 산사태 예측 모델링에 영향을 미칠 수 있음을 시사한다. 향후 이 연구의 결과는 현지 수문모니터링과 함께 수종별 뿌리 점착력의 특징 및 변화를 고려하여 검증되어야 할 것이다.

• ETRI Journal
• /
• 제43권3호
• /
• pp.511-523
• /
• 2021

The World Health Organization provides guidelines for managing the particulate matter (PM) level because a higher PM level represents a threat to human health. To manage the PM level, a procedure for measuring the PM value is first needed. We use a PM sensor that collects the PM level by laser-based light scattering (LLS) method because it is more cost effective than a beta attenuation monitor-based sensor or tapered element oscillating microbalance-based sensor. However, an LLS-based sensor has a higher probability of malfunctioning than the higher cost sensors. In this paper, we regard the overall malfunctioning, including strange value collection or missing collection data as anomalies, and we aim to detect anomalies for the maintenance of PM measuring sensors. We propose a novel architecture for solving the above aim that we call the hypothesis pruning generative adversarial network (HP-GAN). Through comparative experiments, we achieve AUROC and AUPRC values of 0.948 and 0.967, respectively, in the detection of anomalies in LLS-based PM measuring sensors. We conclude that our HP-GAN is a cutting-edge model for anomaly detection.

• 핵의학기술
• /
• 제13권3호
• /
• pp.24-30
• /
• 2009

국내의 암 발생 변화 추이를 고려해보면, 현재 남성에게 발생하는 전립선암의 비중이 나날이 증가하고 있다. 전립선 암의 경우 림프절이나 뼈에 전이가 되면 90% 이상 사망에 이른다. 따라서 최근 핵의학 분야에서는 전립선 암의 뼈 전이를 보다 정확하게 진단하기 위해 여러 영상기법과 방사성 의약품의 개발이 진행, 소개되고 있다. 본 연구에서는 전립선 암세포의 뼈 전이에 대한 보다 정밀한 영상을 획득하여 진단능을 향상시키는 $^{18}F$-Fluoride PET/CT의 유용성을 알아보고자 한다. 본 연구는 총 33명의 전립선암으로 진단된 환자로서, 평균 $67.8{\pm}10.2$세의 남자를 대상으로 하였다. 동일한 환자에게 각각 전신 뼈 검사(Whole Body Bone Scan; WBBS)와 뼈 양전자 방출 컴퓨터 단층 촬영(18F-Fluoride Positron Emission Tomography/Computed Tomography; $^{18}F$-Fluoride PET/CT)을 시행하였으며, 각 검사에서 나타난 병변을 확인하여 ROC곡선(Receiver Operating Characteristic Curve)을 통해 AUROC (The Area Under an ROC)분석하였다. 또한 두 검사의 민감도와 특이도, 양성예측률과 음성예측률 그리고 두 검사의 정확도를 비교 분석하였다. 총 33명 중 22명의 환자(66.6%)에게서 뼈 전이가 있었으며, $^{18}F$-Fluoride PET/CT에서 모두 우수한 결과치가 나타났다. WBBS의 민감도는 63.6%, 특이도는 81.8%였고, $^{18}F$-Fluoride PET/CT에서는 민감도 100%, 특이도 90.9%로 나타났으며, AUROC를 비교한 결과 WBBS에서 0.778, $^{18}F$-Fluoride PET/CT는 0.942로 분석되었다. $^{18}F$-Fluoride PET/CT는 WBBS에 비하여 높은 민감도와 특이도를 보였으며, 3D 영상 구현으로 암세포의 뼈에 대한 전이성 병소를 더욱 정확하고 명확하게 발견할 수 있었다. 또한 WBBS의 장시간 검사에 비하여 $^{18}F$-Fluoride PET/CT는 상대적으로 단시간에 뼈의 생리적인 변화에 대한 정확한 영상을 얻을 수 있어 환자의 불편함을 최소화 할 수 있었다. 고가 검사비용에 대한 부담을 줄일 수 있다면 $^{18}F$-Fluoride PET/CT를 통해 더욱 정확한 영상을 제공하여 임상에서의 진단능을 향상 시킬 수 있을 것으로 사료된다.

• Journal of Yeungnam Medical Science
• /
• 제33권1호
• /
• pp.13-20
• /
• 2016

Background: Prevalence of adenoma in males aged 40-49 years in Korea was higher than expected. The aim of this study was to investigate the prevalence and risk factors of colorectal adenoma in males aged 40-49 years. Methods: Total 1,902 asymptomatic subjects with a mean age of $47.9{\pm}6.7years$, who underwent a screening colonoscopy in a health promotion center of Myongji Hospital from 2010 to 2013 were enrolled in this study. We conducted a case-control study to determine the risk factors for adenoma. The subjects were classified into two groups (adenoma vs. controls). To validate the diagnostic value of carcinoembryonic antigen (CEA) for adenoma, area under the receiver operating characteristic curve (AUROC) was calculated. Results: At least one colorectal adenoma was identified in 385 subjects (20.2%). Among these 385 subjects, 372 subjects were found to have a non-advanced adenoma, 13 subjects had an invasive adenoma. One subject had cancer. Male sex, age, smoking, metabolic syndrome, and elevated CEA level were significantly associated with a colorectal adenoma in univariate analysis. However, metabolic syndrome was not significant in multivariate analysis. In the male group, the AUROC of CEA for colorectal adenoma was 0.600 (0.543 to 0.656) in non-smokers under 50 years of age, and 0.615 (0.540 to 0.690) in smokers under 50 years of age. Conclusion: Male sex, smoking, and high levels of CEA seem to be associated with colorectal adenoma. High levels of CEA and smoking may be diagnostic markers for any colorectal adenoma in Korean males aged 40-49 years.