Yisak Kim
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Yisak Kim has a total of 37 co-authors. Here are the top co-authors by number of shared publications:
- Sung Hye Kong
Department of Internal Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea
3 shared publications - Young-Gon Kim
Seoul National University Hospital
3 shared publications - Byung Woo Kim
Department of Transdisciplinary Medicine, Seoul National University Hospital, Seoul, Korea
2 shared publications - Chang Min Park
Department of Radiology, Seoul National University Hospital, Seoul, Republic of Korea
2 shared publications - Dongheon Lee
Seoul National University College of Medicine, Seoul National University Hospital
2 shared publications - Youmin Shin
Department of Transdisciplinary Medicine, Seoul National University Hospital, Seoul, Korea
2 shared publications
Latest Publications
- Open Access
- 4 November 2026
- Published by Korean College of Rheumatology in Journal of Rheumatic Diseases
Detection of sacroiliitis in spondyloarthritis patients via deep learning from computed tomography scans
Hyonjoung Cho, M.D., Donghoon Kim, M.D., Soyeon Kim, M.S., Yi-sak Kim, Ph.D., Chang Min Park, M.D., Ph.D.. J Rheum Dis -0001;0:. https://doi.org/10.4078/jrd.2026.0035
- 4 November 2026
- Published by Korean College of Rheumatology in Journal of Rheumatic Diseases
Open AccessDetection of sacroiliitis in spondyloarthritis patients via deep learning from computed tomography scans
Hyonjoung Cho, M.D., Donghoon Kim, M.D., Soyeon Kim, M.S., Yi-sak Kim, Ph.D., Chang Min Park, M.D., Ph.D.. J Rheum Dis -0001;0:. https://doi.org/10.4078/jrd.2026.0035
- Open Access
- 12 May 2026
- Published by Elsevier in SSRN Electronic Journal
Mortality Prediction in Elderly Patients with Hip Fracture Using Quantitative Hip Imaging Features
Background: Hip fractures in older adults are associated with high mortality and disability. This study evaluated the association between quantitative 3D CT–derived bone, muscle, and fat features and all-cause mortality in older adults with hip fracture.Methods: We retrospectively studied 748 older adults (mean age, 83.1 ± 10.5 years; 72% women) who underwent preoperative hip CT at a university hospital in Korea between 2016 and 2022. The primary outcome was all-cause mortality, ascertained from national registry data. We extracted 3D features of cortical bone thickness, psoas volume and cross-sectional areas of gluteal muscle and fat, and various femoral geometries using a retrained 3D U-Net segmentation pipeline. Cox proportional hazards regression models were used to evaluate associations between imaging-derived features and mortality.Results: During a mean follow-up of 3.2 years, 287 patients (38.4%) died. Compared with survivors, non-survivors showed thinner cortical bone and reduced muscle measures, including gluteal muscle area and psoas volume. In univariable Cox proportional hazards analyses, thinner cortical bone and reduced muscle measures were associated with increased mortality. In age- and sex-adjusted multivariable models, gluteal muscle and L3-level fat area remained independently associated with all-cause mortality. Sex-stratified analyses demonstrated that muscle-related features were more strongly associated with mortality in men, while gluteal muscle and L3 fat area were the dominant predictors in women.Conclusions: Quantitative 3D CT–derived muscle and fat features are independently associated with mortality following hip fracture. Incorporating these imaging biomarkers into prognostic models may improve risk stratification and support personalized perioperative management.
- 12 May 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessMortality Prediction in Elderly Patients with Hip Fracture Using Quantitative Hip Imaging Features
Background: Hip fractures in older adults are associated with high mortality and disability. This study evaluated the association between quantitative 3D CT–derived bone, muscle, and fat features and all-cause mortality in older adults with hip fracture.Methods: We retrospectively studied 748 older adults (mean age, 83.1 ± 10.5 years; 72% women) who underwent preoperative hip CT at a university hospital in Korea between 2016 and 2022. The primary outcome was all-cause mortality, ascertained from national registry data. We extracted 3D features of cortical bone thickness, psoas volume and cross-sectional areas of gluteal muscle and fat, and various femoral geometries using a retrained 3D U-Net segmentation pipeline. Cox proportional hazards regression models were used to evaluate associations between imaging-derived features and mortality.Results: During a mean follow-up of 3.2 years, 287 patients (38.4%) died. Compared with survivors, non-survivors showed thinner cortical bone and reduced muscle measures, including gluteal muscle area and psoas volume. In univariable Cox proportional hazards analyses, thinner cortical bone and reduced muscle measures were associated with increased mortality. In age- and sex-adjusted multivariable models, gluteal muscle and L3-level fat area remained independently associated with all-cause mortality. Sex-stratified analyses demonstrated that muscle-related features were more strongly associated with mortality in men, while gluteal muscle and L3 fat area were the dominant predictors in women.Conclusions: Quantitative 3D CT–derived muscle and fat features are independently associated with mortality following hip fracture. Incorporating these imaging biomarkers into prognostic models may improve risk stratification and support personalized perioperative management.
- Open Access
- 16 September 2025
- Published by PeerJ in PeerJ Computer Science
- Vol. 11, e3161
Quantification of left ventricular mass in multiple views of echocardiograms using model-agnostic meta learning in a few-shot setting
Background: Reliable measurement of left ventricular mass (LVM) in echocardiography is essential for early detection of left ventricular dysfunction, coronary artery disease, and arrhythmia risk, yet growing patient volumes have created critical shortage of experts in echocardiography. Recent deep learning approaches reduce inter‐operator variability but require large, fully labeled datasets for each standard view—an impractical demand in many clinical settings. Methods: To overcome these limitations, we propose a heatmap-based point-estimation segmentation model trained via model-agnostic meta-learning (MAML) for few-shot LVM quantification across multiple echocardiographic views. Our framework adapts rapidly to new views by learning a shared representation and view-specific head performing K inner-loop updates, and then meta-updating in the outer loop. We used the EchoNet-LVH dataset for the PLAX view, the TMED-2 dataset for the PSAX view and the CAMUS dataset for both the apical 2-chamber and apical 4-chamber views under 1-, 5-, and 10-shot scenarios. Results: As a result, the proposed MAML methods demonstrated comparable performance using mean distance error, mean angle error, successful distance error and spatial angular similarity in a few-shot setting compared to models trained with larger labeled datasets for each view of the echocardiogram.
- 16 September 2025
- Published by PeerJ in PeerJ Computer Science
- Vol. 11, e3161
Open AccessQuantification of left ventricular mass in multiple views of echocardiograms using model-agnostic meta learning in a few-shot setting
Background: Reliable measurement of left ventricular mass (LVM) in echocardiography is essential for early detection of left ventricular dysfunction, coronary artery disease, and arrhythmia risk, yet growing patient volumes have created critical shortage of experts in echocardiography. Recent deep learning approaches reduce inter‐operator variability but require large, fully labeled datasets for each standard view—an impractical demand in many clinical settings. Methods: To overcome these limitations, we propose a heatmap-based point-estimation segmentation model trained via model-agnostic meta-learning (MAML) for few-shot LVM quantification across multiple echocardiographic views. Our framework adapts rapidly to new views by learning a shared representation and view-specific head performing K inner-loop updates, and then meta-updating in the outer loop. We used the EchoNet-LVH dataset for the PLAX view, the TMED-2 dataset for the PSAX view and the CAMUS dataset for both the apical 2-chamber and apical 4-chamber views under 1-, 5-, and 10-shot scenarios. Results: As a result, the proposed MAML methods demonstrated comparable performance using mean distance error, mean angle error, successful distance error and spatial angular similarity in a few-shot setting compared to models trained with larger labeled datasets for each view of the echocardiogram.
- Open Access
- 1 September 2025
- Published by The Korean Society of Nephrology in Kidney Research and Clinical Practice
- Vol. 44 (5) , 802-813
Conventional machine learning-based prediction models did not outperform the International IgA Nephropathy Prediction Tool
Background: Immunoglobulin A nephropathy (IgAN) is a major cause of end-stage kidney disease (ESKD). The International IgA Nephropathy Prediction Tool (IIgAN-PT) predicts IgAN prognosis, but improvement in the prediction performance using machine learning (ML)-based methods is needed. Methods: We analyzed 4,425 biopsy-confirmed patients with IgAN and ≥6 months of follow-up from nine tertiary university hospitals in Korea. The study population was divided into development and validation cohorts. Using the collected 87 clinicodemographic and pathological variables, ML-based prediction models for ESKD or estimated glomerular filtration rate were constructed: 1) the conventional CatBoost model, 2) the optimized CatBoost model with Cox proportional hazards, 3) the deep Cox proportional hazards model, and 4) the deep Cox mixture model. The area under the curve (AUC) and calibration plots were used to investigate the discriminative and calibration performance of the models, which were then compared with those of the IIgAN-PT full model. Results: The full model showed excellent performance (AUC [95% confidence interval] for 5-year outcome, 0.896 [0.8530.940]), with acceptable calibration results. The ML-based models showed good performance in predicting adverse kidney outcomes and revealed acceptable discrimination performance in the external validation (AUC [95% confidence interval] for the 5-year outcome: 1) 0.829 [0.791–0.866]; 2) 0.847 [0.804–0.890]; 3) 0.823 [0.784–0.862]; and 4) 0.832 [0.794–0.870]), although they underestimated the external validation cohort risks. With the validation data, the overall performance of the IIgAN-PT was non-inferior to that of the ML-based model. Conclusions: Our ML-based models showed good performance in predicting adverse kidney outcomes in patients with IgAN but they did not outperform the IIgAN-PT.
- 1 September 2025
- Published by The Korean Society of Nephrology in Kidney Research and Clinical Practice
- Vol. 44 (5) , 802-813
Open AccessConventional machine learning-based prediction models did not outperform the International IgA Nephropathy Prediction Tool
Background: Immunoglobulin A nephropathy (IgAN) is a major cause of end-stage kidney disease (ESKD). The International IgA Nephropathy Prediction Tool (IIgAN-PT) predicts IgAN prognosis, but improvement in the prediction performance using machine learning (ML)-based methods is needed. Methods: We analyzed 4,425 biopsy-confirmed patients with IgAN and ≥6 months of follow-up from nine tertiary university hospitals in Korea. The study population was divided into development and validation cohorts. Using the collected 87 clinicodemographic and pathological variables, ML-based prediction models for ESKD or estimated glomerular filtration rate were constructed: 1) the conventional CatBoost model, 2) the optimized CatBoost model with Cox proportional hazards, 3) the deep Cox proportional hazards model, and 4) the deep Cox mixture model. The area under the curve (AUC) and calibration plots were used to investigate the discriminative and calibration performance of the models, which were then compared with those of the IIgAN-PT full model. Results: The full model showed excellent performance (AUC [95% confidence interval] for 5-year outcome, 0.896 [0.8530.940]), with acceptable calibration results. The ML-based models showed good performance in predicting adverse kidney outcomes and revealed acceptable discrimination performance in the external validation (AUC [95% confidence interval] for the 5-year outcome: 1) 0.829 [0.791–0.866]; 2) 0.847 [0.804–0.890]; 3) 0.823 [0.784–0.862]; and 4) 0.832 [0.794–0.870]), although they underestimated the external validation cohort risks. With the validation data, the overall performance of the IIgAN-PT was non-inferior to that of the ML-based model. Conclusions: Our ML-based models showed good performance in predicting adverse kidney outcomes in patients with IgAN but they did not outperform the IIgAN-PT.
- 1 January 2024
- Published by Radiological Society of North America (RSNA) in Radiology
- Vol. 310 (1) , e230614
A CT-based Deep Learning Model for Predicting Subsequent Fracture Risk in Patients with Hip Fracture
Background Patients have the highest risk of subsequent fractures in the first few years after an initial fracture, yet models to predict short-term subsequent risk have not been developed. Purpose To develop and validate a deep learning prediction model for subsequent fracture risk using digitally reconstructed radiographs from hip CT in patients with recent hip fractures. Materials and Methods This retrospective study included adult patients who underwent three-dimensional hip CT due to a fracture from January 2004 to December 2020. Two-dimensional frontal, lateral, and axial digitally reconstructed radiographs were generated and assembled to construct an ensemble model. DenseNet modules were used to calculate risk probability based on extracted image features and fracture-free probability plots were output. Model performance was assessed using the C index and area under the receiver operating characteristic curve (AUC) and compared with other models using the paired t test. Results The training and validation set included 1012 patients (mean age, 74.5 years ± 13.3 [SD]; 706 female, 113 subsequent fracture) and the test set included 468 patients (mean age, 75.9 years ± 14.0; 335 female, 22 subsequent fractures). In the test set, the ensemble model had a higher C index (0.73) for predicting subsequent fractures than that of other image-based models (C index range, 0.59-0.70 for five of six models; P value range, < .001 to < .05). The ensemble model achieved AUCs of 0.74, 0.74, and 0.73 at the 2-, 3-, and 5-year follow-ups, respectively; higher than that of most other image-based models at 2 years (AUC range, 0.57-0.71 for five of six models; P value range, < .001 to < .05) and 3 years (AUC range, 0.55-0.72 for four of six models; P value range, < .001 to < .05). Moreover, the AUCs achieved by the ensemble model were higher than that of a clinical model that included known risk factors (2-, 3-, and 5-year AUCs of 0.58, 0.64, and 0.70, respectively; P < .001 for all). Conclusion In patients with recent hip fractures, the ensemble deep learning model using digital reconstructed radiographs from hip CT showed good performance for predicting subsequent fractures in the short term. © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Li and Jaremko in this issue.
- 1 January 2024
- Published by Radiological Society of North America (RSNA) in Radiology
- Vol. 310 (1) , e230614
A CT-based Deep Learning Model for Predicting Subsequent Fracture Risk in Patients with Hip Fracture
Background Patients have the highest risk of subsequent fractures in the first few years after an initial fracture, yet models to predict short-term subsequent risk have not been developed. Purpose To develop and validate a deep learning prediction model for subsequent fracture risk using digitally reconstructed radiographs from hip CT in patients with recent hip fractures. Materials and Methods This retrospective study included adult patients who underwent three-dimensional hip CT due to a fracture from January 2004 to December 2020. Two-dimensional frontal, lateral, and axial digitally reconstructed radiographs were generated and assembled to construct an ensemble model. DenseNet modules were used to calculate risk probability based on extracted image features and fracture-free probability plots were output. Model performance was assessed using the C index and area under the receiver operating characteristic curve (AUC) and compared with other models using the paired t test. Results The training and validation set included 1012 patients (mean age, 74.5 years ± 13.3 [SD]; 706 female, 113 subsequent fracture) and the test set included 468 patients (mean age, 75.9 years ± 14.0; 335 female, 22 subsequent fractures). In the test set, the ensemble model had a higher C index (0.73) for predicting subsequent fractures than that of other image-based models (C index range, 0.59-0.70 for five of six models; P value range, < .001 to < .05). The ensemble model achieved AUCs of 0.74, 0.74, and 0.73 at the 2-, 3-, and 5-year follow-ups, respectively; higher than that of most other image-based models at 2 years (AUC range, 0.57-0.71 for five of six models; P value range, < .001 to < .05) and 3 years (AUC range, 0.55-0.72 for four of six models; P value range, < .001 to < .05). Moreover, the AUCs achieved by the ensemble model were higher than that of a clinical model that included known risk factors (2-, 3-, and 5-year AUCs of 0.58, 0.64, and 0.70, respectively; P < .001 for all). Conclusion In patients with recent hip fractures, the ensemble deep learning model using digital reconstructed radiographs from hip CT showed good performance for predicting subsequent fractures in the short term. © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Li and Jaremko in this issue.