James Wei Luen Yip
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
James Wei Luen Yip has a total of 142 co-authors. Here are the top co-authors by number of shared publications:
- Ching-Hui Sia
Department of Cardiology, National University Heart Centre Singapore, Singapore
6 shared publications - Ping Chai
Department of Medicine, Yong Loo Lin School of Medicine, National University of Singapore, Singapore; Department of Cardiology, National University Heart Centre Singapore, Singapore
5 shared publications - Mark Yan Yee Chan
National University Health System
4 shared publications - Tiong-Cheng Yeo
Department of Cardiology, National University Heart Centre Singapore, Singapore 119228, Singapore
4 shared publications - Andrew Fu Wah Ho
SingHealth Duke‐NUS Emergency Medicine Academic Clinical Programme Singapore
3 shared publications - Beng Chin Ooi
Zhejiang University, Hangzhou, China
3 shared publications
Latest Publications
- 16 April 2026
- Published by Springer Nature
- p. 703-719
Adult Congenital Heart Disease
- Open Access
- 7 October 2025
- Published by Wolters Kluwer Health in Journal of the American Heart Association
- Vol. 14 (19) , e038727
Inpatient‐Onset Versus Outpatient‐Onset ST‐Segment–Elevation Myocardial Infarction in Patients With Percutaneous Coronary Intervention: A National Registry Study
Background: Compared with patients who develop ST‐segment–elevation myocardial infarction (STEMI) outside the hospital and present at the emergency department (outpatient‐onset), patients with inpatient‐onset STEMI may paradoxically experience a poorer prognosis due to underlying disease, despite the apparent immediate access to acute health care services. Methods: Patients with first‐onset STEMI who underwent percutaneous coronary intervention ( PCI) in the Singapore Myocardial Infarction Registry (2007–2020) were included. For patients with inpatient‐onset STEMI , the recognition‐to‐balloon time was measured from symptom onset to PCI . For outpatient‐onset STEMI , the recognition‐to‐balloon time was defined as the sum of the symptom‐to‐door and door‐to‐balloon time. Logistic regression was used to identify factors associated with delayed PCI after inpatient‐onset STEMI . Cox regression was used to assess 30‐day, 1‐year, 5‐year, and 10‐year all‐cause and cardiovascular mortality. Results: A total of 19 149 patients, 17 659 (92.2%) outpatient‐onset and 1595 (7.8%) inpatient‐onset, were included. The median follow‐up duration was 6.91 years. Patients with inpatient‐onset STEMI were older, more frequently women, nonsmokers, more likely to have comorbidities, less likely to present with typical STEMI symptoms, and more likely to experience delays in PCI than patients with outpatient‐onset STEMI within the hospital setting. The independent predictors of delayed PCI for inpatient‐onset STEMI were age ≥65 years, diabetes, breathlessness at presentation, and Killip class III. After adjustment for differences in baseline and clinical characteristics, inpatient‐onset STEMI was associated with significantly greater 1‐year, 5‐year, and 10‐year all‐cause mortality (1‐year hazard ratio [HR], 1.27 [95% CI, 1.13–1.43]; 5‐year HR, 1.27 [95% CI, 1.13–1.43]). There was no difference in 30‐day all‐cause or short‐/long‐term cardiovascular mortality. Conclusions: Inpatient‐onset STEMI was linked to an increased long‐term risk of all‐cause mortality compared with outpatient‐onset STEMI despite a shorter recognition‐to‐balloon time.
- 7 October 2025
- Published by Wolters Kluwer Health in Journal of the American Heart Association
- Vol. 14 (19) , e038727
Open AccessInpatient‐Onset Versus Outpatient‐Onset ST‐Segment–Elevation Myocardial Infarction in Patients With Percutaneous Coronary Intervention: A National Registry Study
Background: Compared with patients who develop ST‐segment–elevation myocardial infarction (STEMI) outside the hospital and present at the emergency department (outpatient‐onset), patients with inpatient‐onset STEMI may paradoxically experience a poorer prognosis due to underlying disease, despite the apparent immediate access to acute health care services. Methods: Patients with first‐onset STEMI who underwent percutaneous coronary intervention ( PCI) in the Singapore Myocardial Infarction Registry (2007–2020) were included. For patients with inpatient‐onset STEMI , the recognition‐to‐balloon time was measured from symptom onset to PCI . For outpatient‐onset STEMI , the recognition‐to‐balloon time was defined as the sum of the symptom‐to‐door and door‐to‐balloon time. Logistic regression was used to identify factors associated with delayed PCI after inpatient‐onset STEMI . Cox regression was used to assess 30‐day, 1‐year, 5‐year, and 10‐year all‐cause and cardiovascular mortality. Results: A total of 19 149 patients, 17 659 (92.2%) outpatient‐onset and 1595 (7.8%) inpatient‐onset, were included. The median follow‐up duration was 6.91 years. Patients with inpatient‐onset STEMI were older, more frequently women, nonsmokers, more likely to have comorbidities, less likely to present with typical STEMI symptoms, and more likely to experience delays in PCI than patients with outpatient‐onset STEMI within the hospital setting. The independent predictors of delayed PCI for inpatient‐onset STEMI were age ≥65 years, diabetes, breathlessness at presentation, and Killip class III. After adjustment for differences in baseline and clinical characteristics, inpatient‐onset STEMI was associated with significantly greater 1‐year, 5‐year, and 10‐year all‐cause mortality (1‐year hazard ratio [HR], 1.27 [95% CI, 1.13–1.43]; 5‐year HR, 1.27 [95% CI, 1.13–1.43]). There was no difference in 30‐day all‐cause or short‐/long‐term cardiovascular mortality. Conclusions: Inpatient‐onset STEMI was linked to an increased long‐term risk of all‐cause mortality compared with outpatient‐onset STEMI despite a shorter recognition‐to‐balloon time.
- Open Access
- 8 July 2025
- Published by MDPI AG in Journal of Cardiovascular Development and Disease
- Vol. 12 (7) , 261
The Prognostic Implication of Left Atrial Strain Parameters with Conventional Left Atrial Parameters for the Prediction of Adverse Outcomes in Asian Patients with Hypertrophic Cardiomyopathy—An Echocardiographic Study
Background/Objectives: Left atrial function can be a tool for risk stratification for hypertrophic cardiomyopathy (HCM). Over the past decade, there has been growing interest in the application of strain analysis for earlier and more accurate prediction of cardiovascular disease prognosis. This study aimed to investigate the performance of left atrial strain analysis compared to conventional left atrial measures in predicting clinical outcomes in Asian patients with HCM. Methods and Results: This was a retrospective study involving 291 patients diagnosed with HCM between 2010 and 2017. Left atrial volumes were assessed using the method of discs in orthogonal plans at both end diastole and end systole. Left atrial (LA) strain was obtained using a post-hoc analysis with TOMTEC software. We tested the various left atrial parameters against outcomes of (1) heart failure hospitalization and (2) event-free survival from a composite of adverse events, including all-cause mortality, ventricular tachycardia (VT)/ventricular fibrillation (VF) events, appropriate device therapy if an implantable cardioverter defibrillator (ICD) was implanted, stroke, and heart failure hospitalization. The patients had a mean age of 59.0 ± 16.7 years with a male preponderance (71.2%). The cumulative event-free survival over a follow-up of 3.9 ± 2.7 years was 55.2% for patients with an abnormal LA strain versus 82.4% for patients without one (p < 0.001). Multivariable Cox regression analyses were performed separately for each LA parameter, adjusting for age, sex, LV mass index, LV ejection fraction (EF), E/e’, the presence of LV outflow tract (LVOT) obstruction at rest, and atrial fibrillation. An analysis showed that all parameters except for LAEF demonstrated an independent association with heart failure hospitalization. Left atrial strain outperformed the rest of the parameters by demonstrating an association with a composite of adverse events. Conclusions: In Asian patients with HCM, measures of left atrial strain were independently associated with heart failure hospitalization and a composite of adverse outcomes. Left atrial strain may be used as a tool to predict adverse outcomes in patients with HCM.
- 8 July 2025
- Published by MDPI AG in Journal of Cardiovascular Development and Disease
- Vol. 12 (7) , 261
Open AccessThe Prognostic Implication of Left Atrial Strain Parameters with Conventional Left Atrial Parameters for the Prediction of Adverse Outcomes in Asian Patients with Hypertrophic Cardiomyopathy—An Echocardiographic Study
Background/Objectives: Left atrial function can be a tool for risk stratification for hypertrophic cardiomyopathy (HCM). Over the past decade, there has been growing interest in the application of strain analysis for earlier and more accurate prediction of cardiovascular disease prognosis. This study aimed to investigate the performance of left atrial strain analysis compared to conventional left atrial measures in predicting clinical outcomes in Asian patients with HCM. Methods and Results: This was a retrospective study involving 291 patients diagnosed with HCM between 2010 and 2017. Left atrial volumes were assessed using the method of discs in orthogonal plans at both end diastole and end systole. Left atrial (LA) strain was obtained using a post-hoc analysis with TOMTEC software. We tested the various left atrial parameters against outcomes of (1) heart failure hospitalization and (2) event-free survival from a composite of adverse events, including all-cause mortality, ventricular tachycardia (VT)/ventricular fibrillation (VF) events, appropriate device therapy if an implantable cardioverter defibrillator (ICD) was implanted, stroke, and heart failure hospitalization. The patients had a mean age of 59.0 ± 16.7 years with a male preponderance (71.2%). The cumulative event-free survival over a follow-up of 3.9 ± 2.7 years was 55.2% for patients with an abnormal LA strain versus 82.4% for patients without one (p < 0.001). Multivariable Cox regression analyses were performed separately for each LA parameter, adjusting for age, sex, LV mass index, LV ejection fraction (EF), E/e’, the presence of LV outflow tract (LVOT) obstruction at rest, and atrial fibrillation. An analysis showed that all parameters except for LAEF demonstrated an independent association with heart failure hospitalization. Left atrial strain outperformed the rest of the parameters by demonstrating an association with a composite of adverse events. Conclusions: In Asian patients with HCM, measures of left atrial strain were independently associated with heart failure hospitalization and a composite of adverse outcomes. Left atrial strain may be used as a tool to predict adverse outcomes in patients with HCM.
- Open Access
- 1 October 2024
- Published by Wolters Kluwer Health in Journal of the American Heart Association
- Vol. 13 (19) , e035486
Circulating Plasma Proteins in Aortic Stenosis: Associations With Severity, Myocardial Response, and Clinical Outcomes
Background: Echocardiographic indexes of aortic stenosis may not comprehensively reflect disease morbidity. Plasma proteomic profiling may add prognostic value in these patients. Methods and Results: Proximity extension assays (Olink) of 183 circulating cardiovascular and inflammatory proteins were performed in a prospective follow‐up study of 122 asymptomatic/minimally symptomatic patients (mean±SD age, 69.1±10.9 years; 61% men) with moderate to severe aortic stenosis and preserved left ventricular ejection fraction. Protein signatures of higher‐risk echocardiographic subgroups were determined. Associations of proteins with the primary composite outcome (heart failure hospitalization, progression to New York Heart Association class III‐IV, or all‐cause mortality) were evaluated using competing risk analyses, with aortic valve replacement being the competing risk. Network analysis unveiled mutually exclusive communities of proteins and echocardiographic parameters, connected only through NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide). Members of the tumor necrosis factor receptor superfamily (TNFRSF1A, TNFRSF1B, and TNFRSF14), and trefoil factor‐3 were major hub proteins among the circulating biomarkers. Left ventricular global longitudinal strain >−15% was associated with higher levels of proteins, primarily of inflammation and immune regulation, whereas aortic valve area 15, and left atrial reservoir strain q <0.05). Conclusions: Plasma proteins involved in inflammation and immune regulation were differentially expressed in patients with aortic stenosis with reduced left ventricular global longitudinal strain, and associated with adverse clinical outcomes. Their incorporation into aortic stenosis risk stratification warrants further assessment.
- 1 October 2024
- Published by Wolters Kluwer Health in Journal of the American Heart Association
- Vol. 13 (19) , e035486
Open AccessCirculating Plasma Proteins in Aortic Stenosis: Associations With Severity, Myocardial Response, and Clinical Outcomes
Background: Echocardiographic indexes of aortic stenosis may not comprehensively reflect disease morbidity. Plasma proteomic profiling may add prognostic value in these patients. Methods and Results: Proximity extension assays (Olink) of 183 circulating cardiovascular and inflammatory proteins were performed in a prospective follow‐up study of 122 asymptomatic/minimally symptomatic patients (mean±SD age, 69.1±10.9 years; 61% men) with moderate to severe aortic stenosis and preserved left ventricular ejection fraction. Protein signatures of higher‐risk echocardiographic subgroups were determined. Associations of proteins with the primary composite outcome (heart failure hospitalization, progression to New York Heart Association class III‐IV, or all‐cause mortality) were evaluated using competing risk analyses, with aortic valve replacement being the competing risk. Network analysis unveiled mutually exclusive communities of proteins and echocardiographic parameters, connected only through NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide). Members of the tumor necrosis factor receptor superfamily (TNFRSF1A, TNFRSF1B, and TNFRSF14), and trefoil factor‐3 were major hub proteins among the circulating biomarkers. Left ventricular global longitudinal strain >−15% was associated with higher levels of proteins, primarily of inflammation and immune regulation, whereas aortic valve area 15, and left atrial reservoir strain q <0.05). Conclusions: Plasma proteins involved in inflammation and immune regulation were differentially expressed in patients with aortic stenosis with reduced left ventricular global longitudinal strain, and associated with adverse clinical outcomes. Their incorporation into aortic stenosis risk stratification warrants further assessment.
- 1 September 2024
- Published by Wolters Kluwer Health in Stroke
- Vol. 55 (9) , 2221-2230
Incidence and Outcomes of Cardiocerebral Infarction: A Cohort Study of 2 National Population-Based Registries
BACKGROUND: Cardiocerebral infarction (CCI), which is concomitant with acute myocardial infarction (AMI) and acute ischemic stroke (AIS), is a rare but severe presentation. However, there are few data on CCI, and the treatment options are uncertain. We investigated the characteristics and outcomes of CCI compared with AMI or AIS alone. METHODS: We performed a retrospective cohort study of 120 531 patients with AMI and AIS from the national stroke and AMI registries in Singapore. Patients were categorized into AMI only, AIS only, synchronous CCI (same-day), and metachronous CCI (within 1 week). The primary outcome was all-cause mortality, and the secondary outcome was cardiovascular mortality. The mortality risks were compared using Cox regression. Multivariable models were adjusted for baseline demographics, clinical variables, and treatment for AMI or AIS. RESULTS: Of 127 919 patients identified, 120 531 (94.2%) were included; 74 219 (61.6%) patients had AMI only, 44 721 (37.1%) had AIS only, 625 (0.5%) had synchronous CCI, and 966 (0.8%) had metachronous CCI. The mean age was 67.7 (SD, 14.0) years. Synchronous and metachronous CCI had a higher risk of 30-day mortality (synchronous: adjusted HR [aHR], 2.41 [95% CI, 1.77–3.28]; metachronous: aHR, 2.80 [95% CI, 2.11–3.73]) than AMI only and AIS only (synchronous: aHR, 2.90 [95% CI, 1.87–4.51]; metachronous: aHR, 4.36 [95% CI, 3.03–6.27]). The risk of cardiovascular mortality was higher in synchronous and metachronous CCI than AMI (synchronous: aHR, 3.03 [95% CI, 2.15–4.28]; metachronous: aHR, 3.41 [95% CI, 2.50–4.65]) or AIS only (synchronous: aHR, 2.58 [95% CI, 1.52–4.36]; metachronous: aHR, 4.52 [95% CI, 2.95–6.92]). In synchronous CCI, AMI was less likely to be managed with PCI and secondary prevention medications ( P <0.001) compared with AMI only. CONCLUSIONS: Synchronous CCI occurred in 1 in 200 cases of AIS and AMI. Synchronous and metachronous CCI had higher mortality than AMI or AIS alone.
- 1 September 2024
- Published by Wolters Kluwer Health in Stroke
- Vol. 55 (9) , 2221-2230
Incidence and Outcomes of Cardiocerebral Infarction: A Cohort Study of 2 National Population-Based Registries
BACKGROUND: Cardiocerebral infarction (CCI), which is concomitant with acute myocardial infarction (AMI) and acute ischemic stroke (AIS), is a rare but severe presentation. However, there are few data on CCI, and the treatment options are uncertain. We investigated the characteristics and outcomes of CCI compared with AMI or AIS alone. METHODS: We performed a retrospective cohort study of 120 531 patients with AMI and AIS from the national stroke and AMI registries in Singapore. Patients were categorized into AMI only, AIS only, synchronous CCI (same-day), and metachronous CCI (within 1 week). The primary outcome was all-cause mortality, and the secondary outcome was cardiovascular mortality. The mortality risks were compared using Cox regression. Multivariable models were adjusted for baseline demographics, clinical variables, and treatment for AMI or AIS. RESULTS: Of 127 919 patients identified, 120 531 (94.2%) were included; 74 219 (61.6%) patients had AMI only, 44 721 (37.1%) had AIS only, 625 (0.5%) had synchronous CCI, and 966 (0.8%) had metachronous CCI. The mean age was 67.7 (SD, 14.0) years. Synchronous and metachronous CCI had a higher risk of 30-day mortality (synchronous: adjusted HR [aHR], 2.41 [95% CI, 1.77–3.28]; metachronous: aHR, 2.80 [95% CI, 2.11–3.73]) than AMI only and AIS only (synchronous: aHR, 2.90 [95% CI, 1.87–4.51]; metachronous: aHR, 4.36 [95% CI, 3.03–6.27]). The risk of cardiovascular mortality was higher in synchronous and metachronous CCI than AMI (synchronous: aHR, 3.03 [95% CI, 2.15–4.28]; metachronous: aHR, 3.41 [95% CI, 2.50–4.65]) or AIS only (synchronous: aHR, 2.58 [95% CI, 1.52–4.36]; metachronous: aHR, 4.52 [95% CI, 2.95–6.92]). In synchronous CCI, AMI was less likely to be managed with PCI and secondary prevention medications ( P <0.001) compared with AMI only. CONCLUSIONS: Synchronous CCI occurred in 1 in 200 cases of AIS and AMI. Synchronous and metachronous CCI had higher mortality than AMI or AIS alone.
- Open Access
- 16 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 36 (12) , 7423-7441
Managing Metaverse Data Tsunami: Actionable Insights
In the metaverse the physical space and the virtual space co-exist, and interact simultaneously. While the physical space is virtually enhanced with information, the virtual space is continuously refreshed with real-time, real-world information. To allow users to process and manipulate information seamlessly between the real and digital spaces, novel technologies must be developed. These include smart interfaces, new augmented realities, and efficient data storage, management, and dissemination techniques. In this paper, we first discuss some promising co-space applications. These applications offer opportunities that neither of the spaces can realize on its own. Then, we further discuss several emerging technologies that empower the construction of metaverse. After that, we discuss comprehensively the data centric challenges. Finally, we discuss and envision what are likely to be required from the database and system perspectives.
- 16 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 36 (12) , 7423-7441
Open AccessManaging Metaverse Data Tsunami: Actionable Insights
In the metaverse the physical space and the virtual space co-exist, and interact simultaneously. While the physical space is virtually enhanced with information, the virtual space is continuously refreshed with real-time, real-world information. To allow users to process and manipulate information seamlessly between the real and digital spaces, novel technologies must be developed. These include smart interfaces, new augmented realities, and efficient data storage, management, and dissemination techniques. In this paper, we first discuss some promising co-space applications. These applications offer opportunities that neither of the spaces can realize on its own. Then, we further discuss several emerging technologies that empower the construction of metaverse. After that, we discuss comprehensively the data centric challenges. Finally, we discuss and envision what are likely to be required from the database and system perspectives.
- Open Access
- 9 November 2023
- Published by Oxford University Press (OUP) in European Heart Journal
- Vol. 44 (Supplement)
Incidence and outcomes of type 2 myocardial infarction: a cohort study of a national population-based registry
Background/introduction: Type 2 myocardial infarction (T2MI), caused by myocardial oxygen supply/demand mismatch without acute thrombotic plaque disruption, makes up 2-25% of reported MI. T2MI may be triggered by conditions such as arrhythmias, anaemia, sepsis, heart failure, but the incidence, outcomes and predictors of poor outcomes post-T2MI requires further study. Aims: In this study, we characterised the incidence, outcomes and predictors of poor outcomes in T2MI in Singapore. Methods: We performed an ambidirectional cohort study of 10720 consecutive T2MI patients from January 2012 to December 2018 from the national Singapore MI Registry (SMIR). The primary outcome was major adverse cardiovascular events (MACE; composite of MI, stroke and mortality) within 30 days and 1 year. The secondary outcomes were all-cause mortality, cardiovascular death and stroke. Results: Of 10,724 patients with T2MI, 3852 (35.9%) had MACE within 30 days and 5947 (51.4%) had MACE within 1 year (Table 1). The incidence of T2MI increased from 20 per 100,000 residents in 2012 to 80 per 100,000 residents in 2017 (Figure 1A). Significant independent predictors of MACE at 30 days on multivariable analysis include age (hazard ratio [HR] 1.01, 95% confidence interval (CI) 1.00-1.01), history of hypertension (HR 1.23, 95% CI 1.06-1.42) or diabetes mellitus (HR 1.32, 95% CI 1.17-1.48), Killip classes 2-4 and left ventricular ejection fraction <50% during hospitalization. Low hemoglobin (Hb) 8-9 (HR 1.42, 95% CI 1.07-1.88) and 10-11 (HR 1.36, 95% CI 1.05-1.76) compared to Hb 14-15, were predictors of MACE at 30 days and Hb <13 showed significant associated with MACE at 1 year. Hb <12 was also significantly associated with all-cause mortality at 1 year, but not cardiovascular death or stroke. Significant predictors of stroke at 1 year included current smoking (HR 1.53, 95% CI 1.04-2.23) and atrial fibrillation (AF) (HR 1.54, 95% CI 1.21-1.97). Troponin levels were not significantly associated with MACE at 30 days or 1 year. Conclusions: The incidence of T2MI has been increasing since 2012, with generally poor outcomes within 1 year. Predictors of poor outcomes include age, history of cardiovascular risk factors, AF, reduced LVEF during hospitalization, and low hemoglobin. A higher target Hb may be needed post-T2MI, and further research is needed to establish this.
- 9 November 2023
- Published by Oxford University Press (OUP) in European Heart Journal
- Vol. 44 (Supplement)
Open AccessIncidence and outcomes of type 2 myocardial infarction: a cohort study of a national population-based registry
Background/introduction: Type 2 myocardial infarction (T2MI), caused by myocardial oxygen supply/demand mismatch without acute thrombotic plaque disruption, makes up 2-25% of reported MI. T2MI may be triggered by conditions such as arrhythmias, anaemia, sepsis, heart failure, but the incidence, outcomes and predictors of poor outcomes post-T2MI requires further study. Aims: In this study, we characterised the incidence, outcomes and predictors of poor outcomes in T2MI in Singapore. Methods: We performed an ambidirectional cohort study of 10720 consecutive T2MI patients from January 2012 to December 2018 from the national Singapore MI Registry (SMIR). The primary outcome was major adverse cardiovascular events (MACE; composite of MI, stroke and mortality) within 30 days and 1 year. The secondary outcomes were all-cause mortality, cardiovascular death and stroke. Results: Of 10,724 patients with T2MI, 3852 (35.9%) had MACE within 30 days and 5947 (51.4%) had MACE within 1 year (Table 1). The incidence of T2MI increased from 20 per 100,000 residents in 2012 to 80 per 100,000 residents in 2017 (Figure 1A). Significant independent predictors of MACE at 30 days on multivariable analysis include age (hazard ratio [HR] 1.01, 95% confidence interval (CI) 1.00-1.01), history of hypertension (HR 1.23, 95% CI 1.06-1.42) or diabetes mellitus (HR 1.32, 95% CI 1.17-1.48), Killip classes 2-4 and left ventricular ejection fraction <50% during hospitalization. Low hemoglobin (Hb) 8-9 (HR 1.42, 95% CI 1.07-1.88) and 10-11 (HR 1.36, 95% CI 1.05-1.76) compared to Hb 14-15, were predictors of MACE at 30 days and Hb <13 showed significant associated with MACE at 1 year. Hb <12 was also significantly associated with all-cause mortality at 1 year, but not cardiovascular death or stroke. Significant predictors of stroke at 1 year included current smoking (HR 1.53, 95% CI 1.04-2.23) and atrial fibrillation (AF) (HR 1.54, 95% CI 1.21-1.97). Troponin levels were not significantly associated with MACE at 30 days or 1 year. Conclusions: The incidence of T2MI has been increasing since 2012, with generally poor outcomes within 1 year. Predictors of poor outcomes include age, history of cardiovascular risk factors, AF, reduced LVEF during hospitalization, and low hemoglobin. A higher target Hb may be needed post-T2MI, and further research is needed to establish this.
- Open Access
- 4 August 2023
- Published by Association for Computing Machinery (ACM)
- p. 5071-5081
ECGGAN: A Framework for Effective and Interpretable Electrocardiogram Anomaly Detection
Heart is the most important organ of the human body, and Electrocardiogram (ECG) is an essential tool for clinical monitoring of heart health and detecting cardiovascular diseases. Automatic detection of ECG anomalies is of great significance and clinical value in healthcare. However, performing automatic anomaly detection for the ECG data is challenging because we not only need to accurately detect the anomalies but also need to provide clinically meaningful interpretation of the results. Existing works on automatic ECG anomaly detection either rely on hand-crafted designs of feature extraction algorithms which are typically too simple to deliver good performance, or deep learning for automatically extracting features, which is not interpretable. In this paper, we propose ECGGAN, a novel reconstruction-based ECG anomaly detection framework. The key idea of ECGGAN is to make full use of the characteristics of ECG with the periodic metadata, namely beat, to learn the universal pattern in ECG from representative normal data. We establish a reconstruction model, taking leads as constraints to capture the unique characteristics of different leads in ECG data, and achieve accurate anomaly detection at ECG-level by combining multiple leads. Experimental results on two real-world datasets and their mixed-set confirm that our method achieves superior performance than baselines in terms of precision, recall, F1-score, and AUC. In addition, ECGGAN can provide clinically meaningful interpretation of results by revealing the extent to which abnormal sites deviate from the normal pattern.
- 4 August 2023
- Published by Association for Computing Machinery (ACM)
- p. 5071-5081
Open AccessECGGAN: A Framework for Effective and Interpretable Electrocardiogram Anomaly Detection
Heart is the most important organ of the human body, and Electrocardiogram (ECG) is an essential tool for clinical monitoring of heart health and detecting cardiovascular diseases. Automatic detection of ECG anomalies is of great significance and clinical value in healthcare. However, performing automatic anomaly detection for the ECG data is challenging because we not only need to accurately detect the anomalies but also need to provide clinically meaningful interpretation of the results. Existing works on automatic ECG anomaly detection either rely on hand-crafted designs of feature extraction algorithms which are typically too simple to deliver good performance, or deep learning for automatically extracting features, which is not interpretable. In this paper, we propose ECGGAN, a novel reconstruction-based ECG anomaly detection framework. The key idea of ECGGAN is to make full use of the characteristics of ECG with the periodic metadata, namely beat, to learn the universal pattern in ECG from representative normal data. We establish a reconstruction model, taking leads as constraints to capture the unique characteristics of different leads in ECG data, and achieve accurate anomaly detection at ECG-level by combining multiple leads. Experimental results on two real-world datasets and their mixed-set confirm that our method achieves superior performance than baselines in terms of precision, recall, F1-score, and AUC. In addition, ECGGAN can provide clinically meaningful interpretation of results by revealing the extent to which abnormal sites deviate from the normal pattern.
- Open Access
- 25 May 2023
- Published by MDPI AG in Journal of Cardiovascular Development and Disease
- Vol. 10 (6) , 231
Left Ventricular Ejection Fraction Association with Acute Ischemic Stroke Outcomes in Patients Undergoing Thrombolysis
(1) Background: Little is known about how left ventricular systolic dysfunction (LVSD) affects functional and clinical outcomes in acute ischemic stroke (AIS) patients undergoing thrombolysis; (2) Methods: A retrospective observational study conducted between 2006 and 2018 included 937 consecutive AIS patients undergoing thrombolysis. LVSD was defined as left ventricular ejection fraction (LVEF) < 50%. Univariate and multivariate binary logistic regression analysis was performed for demographic characteristics. Ordinal shift regression was used for functional modified Rankin Scale (mRS) outcome at 3 months. Survival analysis of mortality, heart failure (HF) admission, myocardial infarction (MI) and stroke/transient ischemic attack (TIA) was evaluated with a Cox-proportional hazards model; (3) Results: LVSD patients in comparison with LVEF ≥ 50% patients accounted for 190 and 747 patients, respectively. LVSD patients had more comorbidities including diabetes mellitus (100 (52.6%) vs. 280 (37.5%), p < 0.001), atrial fibrillation (69 (36.3%) vs. 212 (28.4%), p = 0.033), ischemic heart disease (130 (68.4%) vs. 145 (19.4%), p < 0.001) and HF (150 (78.9%) vs. 46 (6.2%), p < 0.001). LVSD was associated with worse functional mRS outcomes at 3 months (adjusted OR 1.41, 95% CI 1.03–1.92, p = 0.030). Survival analysis identified LVSD to significantly predict all-cause mortality (adjusted HR [aHR] 3.38, 95% CI 1.74–6.54, p < 0.001), subsequent HF admission (aHR 4.23, 95% CI 2.17–8.26, p < 0.001) and MI (aHR 2.49, 95% CI 1.44–4.32, p = 0.001). LVSD did not predict recurrent stroke/TIA (aHR 1.15, 95% CI 0.77–1.72, p = 0.496); (4) Conclusions: LVSD in AIS patients undergoing thrombolysis was associated with increased all-cause mortality, subsequent HF admission, subsequent MI and poorer functional outcomes, highlighting a need to optimize LVEF.
- 25 May 2023
- Published by MDPI AG in Journal of Cardiovascular Development and Disease
- Vol. 10 (6) , 231
Open AccessLeft Ventricular Ejection Fraction Association with Acute Ischemic Stroke Outcomes in Patients Undergoing Thrombolysis
(1) Background: Little is known about how left ventricular systolic dysfunction (LVSD) affects functional and clinical outcomes in acute ischemic stroke (AIS) patients undergoing thrombolysis; (2) Methods: A retrospective observational study conducted between 2006 and 2018 included 937 consecutive AIS patients undergoing thrombolysis. LVSD was defined as left ventricular ejection fraction (LVEF) < 50%. Univariate and multivariate binary logistic regression analysis was performed for demographic characteristics. Ordinal shift regression was used for functional modified Rankin Scale (mRS) outcome at 3 months. Survival analysis of mortality, heart failure (HF) admission, myocardial infarction (MI) and stroke/transient ischemic attack (TIA) was evaluated with a Cox-proportional hazards model; (3) Results: LVSD patients in comparison with LVEF ≥ 50% patients accounted for 190 and 747 patients, respectively. LVSD patients had more comorbidities including diabetes mellitus (100 (52.6%) vs. 280 (37.5%), p < 0.001), atrial fibrillation (69 (36.3%) vs. 212 (28.4%), p = 0.033), ischemic heart disease (130 (68.4%) vs. 145 (19.4%), p < 0.001) and HF (150 (78.9%) vs. 46 (6.2%), p < 0.001). LVSD was associated with worse functional mRS outcomes at 3 months (adjusted OR 1.41, 95% CI 1.03–1.92, p = 0.030). Survival analysis identified LVSD to significantly predict all-cause mortality (adjusted HR [aHR] 3.38, 95% CI 1.74–6.54, p < 0.001), subsequent HF admission (aHR 4.23, 95% CI 2.17–8.26, p < 0.001) and MI (aHR 2.49, 95% CI 1.44–4.32, p = 0.001). LVSD did not predict recurrent stroke/TIA (aHR 1.15, 95% CI 0.77–1.72, p = 0.496); (4) Conclusions: LVSD in AIS patients undergoing thrombolysis was associated with increased all-cause mortality, subsequent HF admission, subsequent MI and poorer functional outcomes, highlighting a need to optimize LVEF.
- Open Access
Toward Cohort Intelligence: A Universal Cohort Representation Learning Framework for Electronic Health Record Analysis
Electronic Health Records (EHR) are generated from clinical routine care recording valuable information of broad patient populations, which provide plentiful opportunities for improving patient management and intervention strategies in clinical practice. To exploit the enormous potential of EHR data, a popular EHR data analysis paradigm in machine learning is EHR representation learning, which first leverages the individual patient's EHR data to learn informative representations by a backbone, and supports diverse health-care downstream tasks grounded on the representations. Unfortunately, such a paradigm fails to access the in-depth analysis of patients' relevance, which is generally known as cohort studies in clinical practice. Specifically, patients in the same cohort tend to share similar characteristics, implying their resemblance in medical conditions such as symptoms or diseases. In this paper, we propose a universal COhort Representation lEarning (CORE) framework to augment EHR utilization by leveraging the fine-grained cohort information among patients. In particular, CORE first develops an explicit patient modeling task based on the prior knowledge of patients' diagnosis codes, which measures the latent relevance among patients to adaptively divide the cohorts for each patient. Based on the constructed cohorts, CORE recodes the pre-extracted EHR data representation from intra- and inter-cohort perspectives, yielding augmented EHR data representation learning. CORE is readily applicable to diverse backbone models, serving as a universal plug-in framework to infuse cohort information into healthcare methods for boosted performance. We conduct an extensive experimental evaluation on two real-world datasets, and the experimental results demonstrate the effectiveness and generalizability of CORE.
Open AccessToward Cohort Intelligence: A Universal Cohort Representation Learning Framework for Electronic Health Record Analysis
Electronic Health Records (EHR) are generated from clinical routine care recording valuable information of broad patient populations, which provide plentiful opportunities for improving patient management and intervention strategies in clinical practice. To exploit the enormous potential of EHR data, a popular EHR data analysis paradigm in machine learning is EHR representation learning, which first leverages the individual patient's EHR data to learn informative representations by a backbone, and supports diverse health-care downstream tasks grounded on the representations. Unfortunately, such a paradigm fails to access the in-depth analysis of patients' relevance, which is generally known as cohort studies in clinical practice. Specifically, patients in the same cohort tend to share similar characteristics, implying their resemblance in medical conditions such as symptoms or diseases. In this paper, we propose a universal COhort Representation lEarning (CORE) framework to augment EHR utilization by leveraging the fine-grained cohort information among patients. In particular, CORE first develops an explicit patient modeling task based on the prior knowledge of patients' diagnosis codes, which measures the latent relevance among patients to adaptively divide the cohorts for each patient. Based on the constructed cohorts, CORE recodes the pre-extracted EHR data representation from intra- and inter-cohort perspectives, yielding augmented EHR data representation learning. CORE is readily applicable to diverse backbone models, serving as a universal plug-in framework to infuse cohort information into healthcare methods for boosted performance. We conduct an extensive experimental evaluation on two real-world datasets, and the experimental results demonstrate the effectiveness and generalizability of CORE.