Sun-Hwa Kim’s research while affiliated with Kyungpook National University and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (60)


Steps of AI-QCA process. (A) DICOM upload page for web-based AI-QCA version. (B) The best frame for QCA is recommended: 42nd frame out of 111 frames in this case. One can play the video by clicking the play button, or switch to other frames by clicking the forward/backward buttons or dragging the slide. When “Analyze” button is clicked, (C) AI-assisted QCA is performed. In this case, the most severe lesion in the mid-left anterior descending artery is shown first. P and D indicate proximal and distal lesion margins, respectively. Lesion length, diameter stenosis (DS), minimum lumen diameter (MLD), proximal reference diameter (D), distal reference diameter (D), and reference diameter (RD) are shown. Estimated lumen diameter along the vessel is shown in the diagram. (D) By clicking “S#1” tab, analysis on diagonal branch is presented. AI-QCA: artificial intelligence-based quantitative coronary angiography; DICOM, Digital Imaging and Communications in Medicine; QCA, quantitative coronary angiography.
Scatter plot analysis between AI-QCA and manual QCA (phase 2). The plot shows strong correlations across all measurements: proximal RD, distal RD, RD, MLD, %DS, and LL. Gray regions indicate 95% confidence intervals for each measurement. AI-QCA, artificial intelligence-based quantitative coronary angiography; %DS, percent diameter stenosis; LL, lesion length; MLD, minimal lumen diameter; QCA, quantitative coronary angiography; RD, reference diameter.
Bland-Altman plot between AI-QCA and manual QCA (phase 2). The points are distributed around the central line (black solid line) and fall within the limits of agreement (red dashed lines), indicating a high concordance between the two methods. AI-QCA, artificial intelligence-based quantitative coronary angiography; QCA, quantitative coronary angiography.
Intraclass correlation coefficient indicating the consistency of lesion parameters between AI-QCA versus manual QCA by different observers in phase 1 pilot study (n = 15).
Correlation coefficients and p values of lesion parameters of AI-QCA and visual assessment with manual QCA calculated using Pearson's method.

+3

Validation of artificial intelligence-based quantitative coronary angiography
  • Article
  • Full-text available

December 2024

·

13 Reads

Do-Hyun Kim

·

Sun-Hwa Kim

·

Hyun-Wook Chu

·

[...]

·

Background Coronary angiography is fundamental for the diagnosis and treatment of coronary artery disease. Manual quantitative coronary angiography (QCA) is accurate and reproducible; however, it is time-consuming and labor-intensive. However, recent advancements in artificial intelligence (AI) have enabled automated and rapid analysis of medical images, addressing the need for real-time quantitative coronary analysis. Aims This study aimed to evaluate the accuracy of AI-based QCA (AI-QCA) compared with that via manual QCA and clinician acceptance. Methods This retrospective, single-center study was conducted in two phases. Phase 1 was a pilot study comparing AI-QCA with manual QCA and visual estimation. It involved 15 patients who underwent coronary angiography at Seoul National University Bundang Hospital between September 2011 and July 2021. Phase 2 included a larger cohort of 762 patients, with 1002 coronary angiograms analyzed between May 2020 and April 2021. Results In phase 1, AI-QCA and manual QCA consistency varied among the observers, with AI-QCA showing superior consistency compared with visual estimation. However, a strong correlation between AI-QCA and manual-QCA was found in phase 2. AI-QCA accurately identified and quantitatively analyzed multiple lesions in the major vessels, providing results comparable with those of manual QCA. Conclusions AI-QCA demonstrated high concordance with manual QCA, offering real-time analysis and reduced workload. Therefore, AI-QCA has the potential to be a valuable tool for diagnosing and treating coronary artery disease, necessitating further studies for clinical validation.

Download


Effective descriptor extraction strategies for correspondence matching in coronary angiography images

August 2024

·

24 Reads

The importance of 3D reconstruction of coronary arteries using multiple coronary angiography (CAG) images has been increasingly recognized in the field of cardiovascular disease management. This process relies on the camera matrix’s optimization, needing correspondence info for identical point positions across two images. Therefore, an automatic method for determining correspondence between two CAG images is highly desirable. Despite this need, there is a paucity of research focusing on image matching in the CAG images. Additionally, standard deep learning image matching techniques often degrade due to unique features and noise in CAG images. This study aims to fill this gap by applying a deep learning-based image matching method specifically tailored for the CAG images. We have improved the structure of our point detector and redesigned loss function to better handle sparse labeling and indistinct local features specific to CAG images. Our method include changes to training loss and introduction of a multi-head descriptor structure leading to an approximate 6% improvement. We anticipate that our work will provide valuable insights into adapting techniques from general domains to more specialized ones like medical imaging and serve as an improved benchmark for future endeavors in X-ray image-based correspondence matching.


Diabetes Primes Neutrophils for Neutrophil Extracellular Trap Formation through Trained Immunity

April 2024

·

74 Reads

·

4 Citations

Neutrophils are primed for neutrophil extracellular trap (NET) formation during diabetes, and excessive NET formation from primed neutrophils compromises wound healing in patients with diabetes. Here, we demonstrate that trained immunity mediates diabetes-induced NET priming in neutrophils. Under diabetic conditions, neutrophils exhibit robust metabolic reprogramming comprising enhanced glycolysis via the pentose phosphate pathway and fatty acid oxidation, which result in the accumulation of acetyl-coenzyme A. Adenosine 5′-triphosphate-citrate lyase-mediated accumulation of acetyl-coenzyme A and histone acetyltransferases further induce the acetylation of lysine residues on histone 3 (AcH3K9, AcH3K14, and AcH3K27) and histone 4 (AcH4K8). The pharmacological inhibition of adenosine 5′-triphosphate-citrate lyase and histone acetyltransferases completely inhibited high-glucose-induced NET priming. The trained immunity of neutrophils was further confirmed in neutrophils isolated from patients with diabetes. Our findings suggest that trained immunity mediates functional changes in neutrophils in diabetic environments, and targeting neutrophil-trained immunity may be a potential therapeutic target for controlling inflammatory complications of diabetes.


Fig. 1. Flow chart. LVEF = left ventricular ejection fraction, PEF = preserved ejection fraction, REF = reduced ejection fraction, HPR = high platelet reactivity, mrEF = mildly reduced ejection fraction.
Fig. 2. Kaplan-meier curves between PEF group and REF group. PEF = preserved ejection fraction, REF = reduced ejection fraction, HR = hazard ratio, CI = confidence interval.
Clinical outcomes according to left ventricular ejection fraction and/or high platelet reactivity
Prognostic Implication of Platelet Reactivity According to Left Ventricular Systolic Dysfunction Status in Patients Treated With Drug-Eluting Stent Implantation: Analysis of the PTRG-DES Consortium

January 2024

·

34 Reads

·

1 Citation

Journal of Korean medical science

Background Coronary artery disease patients undergoing percutaneous coronary intervention (PCI) often exhibit reduced left ventricular ejection fraction (LVEF). However, the impact of LV dysfunction status in conjunction with platelet reactivity on clinical outcomes has not been previously investigated. Methods From the multicenter PTRG-DES (Platelet function and genoType-Related long-term prognosis in DES-treated patients) consortium, the patients were classified as preserved-EF (PEF: LVEF ≥ 50%) and reduced-EF (REF: LVEF< 5 0%) group by echocardiography. Platelet reactivity was measured using VerifyNow P2Y12 assay and high platelet reactivity (HPR) was defined as PRU ≥ 252. The major adverse cardiac and cerebrovascular events (MACCEs) were a composite of death, myocardial infarction, stent thrombosis and stroke at 5 years after PCI. Major bleeding was defined as Bleeding Academic Research Consortium bleeding types 3–5. Results A total of 13,160 patients from PTRG-DES, 9,319 (79.6%) patients with the results of both PRU and LVEF were analyzed. The incidence of MACCE and major bleeding was higher in REF group as compared with PEF group (MACCEs: hazard ratio [HR] 2.17, P < 0.001, 95% confidence interval [CI] 1.85–2.55; major bleeding: HR 1.78, P < 0.001, 95% CI 1.39–2.78). The highest rate of MACCEs was found in patients with REF and HPR, and the difference between the groups was statistically significant (HR 3.14 in REF(+)/HPR(+) vs. PEF(+)/HPR(-) group, P < 0.01, 95% CI 2.51–3.91). The frequency of major bleeding was not associated with the HPR in either group. Conclusion LV dysfunction was associated with an increased incidence of MACCEs and major bleeding in patients who underwent PCI. The HPR status further exhibited significant increase of MACCEs in patients with LV dysfunction in a large, real-world registry. Trial Registration ClinicalTrials.gov Identifier: NCT04734028


Description of the study population
Unadjusted cardiovascular risk profiles according to toothbrushing frequency. (A) Estimated 10-year atherosclerotic cardiovascular risk calculated using pooled cohort equations, (B) high-sensitivity C-reactive protein (hsCRP), and (C) white blood cell (WBC) counts. Multivariable adjustment for the association between toothbrushing frequency and cardiovascular risk profiles (D). The estimated 10-year ASCVD risk was log-transformed. The red box and line indicate a comparison of toothbrushing ≥ 3 vs. 0−1 times/day; blue indicates that of 2 vs. 0−1 times/day. *Adjusted for age, sex, obesity, current smoking, heavy alcohol drinking, and physical activity; **Adjusted for age, sex, obesity, current smoking, heavy alcohol drinking, physical activity, hypertension, diabetes, and dyslipidaemia. Abbreviations: hsCRP, high-sensitivity C-reactive protein; WBC, white blood cell
Association between toothbrushing and cardiovascular risk factors: a cross-sectional study using Korean National Health and Nutrition Examination Survey 2015–2017

January 2024

·

83 Reads

·

8 Citations

BMC Oral Health

Background Previous studies have suggested that frequent toothbrushing is associated with a lower risk of future cardiovascular events. We sought to investigate further the relationship between toothbrushing, cardiovascular risk factors, and lifestyle behaviours. Methods We analysed a cross-sectional survey including 13,761 adults aged 30 years or older without a history of cardiovascular diseases from the Korean National Health and Nutritional Examination Survey. Conventional cardiovascular risk factors (blood pressure, lipid profiles, and fasting glucose), and inflammatory markers (high-sensitivity C-reactive protein [hsCRP], and white blood cell counts [WBC]) were investigated in relation to the frequency of toothbrushing. Results The estimated 10-year atherosclerotic cardiovascular disease (ASCVD) risk, calculated using the pooled cohort equations was 13.7%, 9.1%, and 7.3% for participants who reported toothbrushing 0–1, 2, and ≥ 3 times a day, respectively. Both conventional risk factors and inflammatory markers were significantly associated with frequent toothbrushing. However, after adjusting potential confounding factors such as age, sex, comorbidities, and lifestyle behaviours, only inflammatory markers were remained as significant factors. Conclusions Oral hygiene behaviours are closely linked to cardiovascular risk factors. This study suggests that reduced systemic inflammatory burden may explain the benefit of improved oral hygiene in terms of cardiovascular risk.


Flowchart of dataset creation.
Performance of algorithms in diagnosis.
Ranked importance scores.
Machine learning for predicting hepatitis B or C virus infection in diabetic patients

December 2023

·

57 Reads

·

1 Citation

Highly prevalent hepatitis B and hepatitis C virus (HBV and HCV) infections have been reported among individuals with diabetes. Given the frequently asymptomatic nature of hepatitis and the challenges associated with screening in some vulnerable populations such as diabetes patients, we conducted an investigation into the performance of various machine learning models for the identification of hepatitis in diabetic patients while also evaluating the significance of features. Analyzing NHANES data from 2013 to 2018, machine learning models were evaluated; random forest (RF), support vector machine (SVM), eXtreme Gradient Boosting (XGBoost), and least absolute shrinkage and selection operator (LASSO) along with stacked ensemble model. We performed hyperparameter tuning to improve the performance of the model, and selected important predictors using the best performance model. LASSO showed the highest predictive performance (AUC-ROC = 0.810) rather than other models. Illicit drug use, poverty, and race were highly ranked as predictive factors for developing hepatitis in diabetes patients. Our study demonstrated that a machine-learning-based model performed optimally in the detection of hepatitis among diabetes patients, achieving high performance. Furthermore, models and predictors evaluated from the current study, we expect, could be supportive information for developing screening or treatment methods for hepatitis care in diabetes patients.


Figure 1
Effects of Optimal Medical Therapy in Patients Undergoing PCI: Subgroup Analysis of the POST-PCI Trial

September 2023

·

66 Reads

Coronary artery disease (CAD) is a major global health concern, and percutaneous coronary intervention (PCI) is commonly used for revascularization. Optimal medical therapy (OMT) is recommended to complement PCI. However, whether OMT significantly benefits high-risk CAD patients remains unclear. This study aimed to investigate potential benefits of OMT in high-risk CAD patients who underwent PCI, through a post-hoc subgroup analysis of the POST-PCI trial comparing patients who did and did not receive OMT. The primary endpoint was major cardiovascular events (death from any cause, myocardial infarction, or hospitalization for unstable angina) at 2 years; secondary endpoints included individual components of the primary outcome. The OMT group comprised patients with a higher prevalence of cardiovascular risk factors. However, clinical outcomes did not differ significantly between OMT and non-OMT groups. Additionally, the OMT group had a significantly reduced invasive coronary angiography rate than the non-OMT group (74.0% vs. 89.1%, odds ratio: 0.35, 95% confidence interval: 0.15–0.81, p = 0.014). The non-OMT group tended to have a higher proportion of symptomatic patients undergoing repeat revascularization than the OMT group (81.7% vs. 65.5%, p = 0.070). In conclusion, OMT did not result in significantly different clinical outcomes in high-risk CAD patients post-PCI compared with non-administration of OMT. However, OMT has potential benefits in reducing unnecessary procedures and optimizing patient care. Further studies are needed to better understand OMT benefits in this patient population.


Performance of a Novel CT-Derived Fractional Flow Reserve Measurement to Detect Hemodynamically Significant Coronary Stenosis

August 2023

·

13 Reads

·

2 Citations

Journal of Korean medical science

Background: Fractional flow reserve (FFR) based on computed tomography (CT) has been shown to better identify ischemia-causing coronary stenosis. However, this current technology requires high computational power, which inhibits its widespread implementation in clinical practice. This prospective, multicenter study aimed at validating the diagnostic performance of a novel simple CT based fractional flow reserve (CT-FFR) calculation method in patients with coronary artery disease. Methods: Patients who underwent coronary CT angiography (CCTA) within 90 days and invasive coronary angiography (ICA) were prospectively enrolled. A hemodynamically significant lesion was defined as an FFR ≤ 0.80, and the area under the receiver operating characteristic curve (AUC) was the primary measure. After the planned analysis for the initial algorithm A, we performed another set of exploratory analyses for an improved algorithm B. Results: Of 184 patients who agreed to participate in the study, 151 were finally analyzed. Hemodynamically significant lesions were observed in 79 patients (52.3%). The AUC was 0.71 (95% confidence interval [CI], 0.63-0.80) for CCTA, 0.65 (95% CI, 0.56-0.74) for CT-FFR algorithm A (P = 0.866), and 0.78 (95% CI, 0.70-0.86) for algorithm B (P = 0.112). Diagnostic accuracy was 0.63 (0.55-0.71) for CCTA alone, 0.66 (0.58-0.74) for algorithm A, and 0.76 (0.68-0.82) for algorithm B. Conclusion: This study suggests the feasibility of automated CT-FFR, which can be performed on-site within several hours. However, the diagnostic performance of the current algorithm does not meet the a priori criteria for superiority. Future research is required to improve the accuracy.


Impact of dementia and drug compliance on patients with acute myocardial infarction

July 2023

·

14 Reads

·

1 Citation

Background: In South Korea, the number of people with dementia is rising at a worrisome rate, and many of them also have acute myocardial infarction (AMI), a disease with a high mortality rate. Hypothesis: We speculated that dementia and drug compliance have significant impact on the mortality of patients with AMI. Methods: The study derived data from the National Health Insurance Service-Senior for a retrospective cohort study. The total number of patients diagnosed with AMI for the first time between 2007 and 2013 was 16 835, among whom 2021 had dementia. Medication possession ratio (MPR) was used to assess medication adherence. Results: AMI patients with dementia had unfavorable baseline characteristics; they had significantly higher risk of all-cause mortality (hazard ratio [HR]: 2.49; 95% confidence interval [CI]: 2.34-2.66; p < .001) and lower MPR (aspirin: 21.9% vs. 42.8%; p < .001). AMI patients were stratified by presence of dementia and medication adherence, and the survival rate was the highest among those with no dementia and good adherence, followed by those with no dementia and poor adherence, those with dementia and good adherence, and those with dementia and poor adherence. The multivariable analysis revealed that dementia (HR: 1.64; 95% CI: 1.53-1.75; p < .001) and poor adherence to medication (HR: 1.60; 95% CI: 1.49-1.71; p < .001) had a significant association with all-cause mortality in AMI patients. Conclusions: AMI patients with dementia have a higher mortality rate. Their prognosis is negatively affected by their poorer medication adherence than patients without dementia.


Citations (37)


... Moreover, at the molecular Immune cells require glucose for energy production [102]. Like adipose, muscle, and liver cells, they have insulin receptors (IRs) on their surfaces [103,104]. Insulin, functioning as a glucose-regulating hormone through IRs, also acts as a growth factor and cytokine regulator, thereby influencing immune modulation [105][106][107]. Insulin influences the immune response both indirectly by lowering glucose levels and directly by affecting immune cells, impacting their proliferation and signal transduction [108,109]. ...

Reference:

Diabetes and Osteoarthritis: Exploring the Interactions and Therapeutic Implications of Insulin, Metformin, and GLP-1-Based Interventions
Diabetes Primes Neutrophils for Neutrophil Extracellular Trap Formation through Trained Immunity

... Subsequently, a study conducted by Reichert et al. in 2015 showed that flossing and brushing of interdental spaces were associated with reduced risk for new cardiovascular events among patients with a history of CAD [5]. In another study, improved oral hygiene was associated with reduced systemic inflammatory risk factors for CVD [6]. ...

Association between toothbrushing and cardiovascular risk factors: a cross-sectional study using Korean National Health and Nutrition Examination Survey 2015–2017

BMC Oral Health

... Coronary artery disease (CAD) is one of the major causes of left ventricular systolic dysfunction, progressive decline in cardiac function may occur in patients with CAD even after revascularization, which subsequently leads to heart failure (HF) (1,2). Meanwhile, due to the presence of cardiovascular risk factors, adverse myocardial remodeling and dysfunction appear much earlier than the onset of HF symptoms (3). ...

Prognostic Implication of Platelet Reactivity According to Left Ventricular Systolic Dysfunction Status in Patients Treated With Drug-Eluting Stent Implantation: Analysis of the PTRG-DES Consortium

Journal of Korean medical science

... The outcomes of this study revealed that the random forest model was the best at predicting HCV, with 94.88% accuracy and a recall of 84.52%. Kim et al. [13] employed National Health and Nutrition Examination Survey (NHANES) data from 2013-2018 to build a model predicting HCV and Hepatitis B Virus (HBV) among diabetic patients. Their datasets included 1,396 diabetic samples with mean age of 54.66 years, among which 64 samples were those infected by HBV or HCV and 1,332 records were without HBV or HCV infection. ...

Machine learning for predicting hepatitis B or C virus infection in diabetic patients

... Attached for 11 days, the AT-Patch recorded continuous heart electrical activity and detected AF in 3.4% of participants, as confirmed by two independent cardiologists. The device's reliability was supported by an overall Cohen κ coefficient of 0.840 [19]. In comparison, the Zio® Patch by iRhythm Technologies is a single-use, waterproof, continuous ECG monitor worn for up to 14 days. ...

Diagnostic value of the AT-Patch, a wearable continuous electrocardiogram monitoring device, for new-onset atrial fibrillation in high-risk patients (ESTRENO-AF): a prospective cohort study (Preprint)

Journal of Medical Internet Research

... В корейском исследовании, посвящённом анализу приверженности (по PDC) фиксированных (ФК) и нефиксированных комбинаций (НФК) БРА + статин за период 540 дней превосходство ФК по сокращению смерти от всех причин не было доказано [56]. Корейские учёные также показали, что пациенты с острым ИМ и деменцией имели значительно более высокий риск смертности от всех причин (ОР 2,49; 95 % ДИ 2,34-2,66; p<0,001) и более низкий MPR по аспирину (21,9 против 42,8 %; р<0,001) по сравнению с больными без деменцией [58]. Многофакторный анализ выявил, что деменция (ОР 1,64; 95 % ДИ 1,53-1,75; p<0,001) и плохая приверженность лечению (ОР 1,60; 95 % ДИ 1,49−1,71; p<0,001) имеют существенное значение по влиянию на смертность пациентов с острым ИМ. ...

Impact of dementia and drug compliance on patients with acute myocardial infarction
  • Citing Article
  • July 2023

... In Kim et al. 16 demonstrated a strong correlation between AI-QCA and manual QCA for identifying lumen diameters. Similarly, Moon et al. 21 showed a moderate to strong correlation between AI-QCA and intravascular ultrasound when analyzing coronary lesions with significant stenosis. ...

Accuracy of Artificial Intelligence-Based Automated Quantitative Coronary Angiography Compared to Intravascular Ultrasound: Retrospective Cohort Study. (Preprint)

JMIR Cardio

... The current study evaluated illegal drug injection, poverty, and race as key factors for predicting hepatitis. The results of major predictor variables related to hepatitis prevalence in diabetic patients, as observed in the study by Han et al. 38 , were consistent with the findings of our study. Illegal drug use occurs globally, negatively affects the quality of life of individuals and communities, reduces productivity, and significantly increases the demands on healthcare systems 39 . ...

Hepatitis Risk in Diabetes Compared to Non-Diabetes and Relevant Factors: A Cross-Sectional Study with National Health and Nutrition Examination Survey (NHANES), 2013–2018

... Though these have beneficial attributes compared with BMS, their superiority against new-generation polymer-based DES has yet to be shown. A simultaneous decrease in strut thickness with biodegradable polymers to prepare ultrathin BP DES has a great potential (Kandzari et al., 2022;Yoon et al., 2023). Polylactic acid and its derivatives are the most used BP in the BP-DES (Hassan et al., 2022;Schwartz et al., 2002). ...

BioMatrix Versus Orsiro Stents for Coronary Artery Disease: A Multicenter, Randomized, Open-Label Study
  • Citing Article
  • December 2022

Circulation Cardiovascular Interventions

... Based on the literature, it can be said that the LA and LV strains seem to be suitable for assessing the success of certain drug treatments, e.g., angiotensin-converting enzyme inhibitors [30], angiotensin receptor neprilysin inhibitors [31], and beta-blockers [32]. One may wonder how these deformation parameters could serve as biomarkers in the followup. ...

Reverse Remodeling Assessed by Left Atrial and Ventricular Strain Reflects Treatment Response to Sacubitril/Valsartan
  • Citing Article
  • May 2022

JACC Cardiovascular Imaging