E. L. Sandberg, S. Halvorsen, T. Berge, Jostein Grimsmo, D. Atar, Bjørnar Leangen Grenne, J. Jortveit
Atrial fibrillation (AF) is prevalent, undiagnosed in approximately one-third of cases, and is associated with severe complications. Guidelines recommend screening individuals at increased risk of stroke. This report evaluated the digital recruitment procedure and compliance with the follow-up recommendations in participants with screen-detected AF in the Norwegian Atrial Fibrillation self-screening pilot study. Norwegians ≥65 years were invited through Facebooks posts, web pages and newspapers to participate in the study. Targeted Facebook posts promoted over 11 days reached 84,208 users, and 10,582 visitors to the study homepage. This accounted for 51% of the total homepage visitors (n=20,704). A total of 2,118 (10%) of the homepage visitors provided digital consent to participate after they met the inclusion criteria. The mean (SD) age of the participants was 70 (4) years, and the majority (n=1,569 (74%)) were women. A total of 1,849 (87%) participants completed the ECG self-screening test, identifying AF in 41 (2.2%) individuals. Of these, 39 (95%) participants consulted a general practitioner (GP), and 34 (83%) participants initiated anticoagulation therapy. Digital recruitment and inclusion in digital AF screening with a high rate of initiation of anticoagulation therapy in AF positive screening cases are feasible. However, digital recruitment and inclusion may introduce selection bias with regard to age and gender. Larger studies are needed to determine the efficacy and cost-effectiveness of a fully digital AF screening.
{"title":"Digital recruitment and compliance to treatment recommendations in the Norwegian Atrial Fibrillation self-screening pilot study","authors":"E. L. Sandberg, S. Halvorsen, T. Berge, Jostein Grimsmo, D. Atar, Bjørnar Leangen Grenne, J. Jortveit","doi":"10.1093/ehjdh/ztae026","DOIUrl":"https://doi.org/10.1093/ehjdh/ztae026","url":null,"abstract":"\u0000 \u0000 \u0000 Atrial fibrillation (AF) is prevalent, undiagnosed in approximately one-third of cases, and is associated with severe complications. Guidelines recommend screening individuals at increased risk of stroke. This report evaluated the digital recruitment procedure and compliance with the follow-up recommendations in participants with screen-detected AF in the Norwegian Atrial Fibrillation self-screening pilot study.\u0000 \u0000 \u0000 \u0000 Norwegians ≥65 years were invited through Facebooks posts, web pages and newspapers to participate in the study. Targeted Facebook posts promoted over 11 days reached 84,208 users, and 10,582 visitors to the study homepage. This accounted for 51% of the total homepage visitors (n=20,704). A total of 2,118 (10%) of the homepage visitors provided digital consent to participate after they met the inclusion criteria. The mean (SD) age of the participants was 70 (4) years, and the majority (n=1,569 (74%)) were women. A total of 1,849 (87%) participants completed the ECG self-screening test, identifying AF in 41 (2.2%) individuals. Of these, 39 (95%) participants consulted a general practitioner (GP), and 34 (83%) participants initiated anticoagulation therapy.\u0000 \u0000 \u0000 \u0000 Digital recruitment and inclusion in digital AF screening with a high rate of initiation of anticoagulation therapy in AF positive screening cases are feasible. However, digital recruitment and inclusion may introduce selection bias with regard to age and gender. Larger studies are needed to determine the efficacy and cost-effectiveness of a fully digital AF screening.\u0000","PeriodicalId":508387,"journal":{"name":"European Heart Journal - Digital Health","volume":"9 4","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140727286","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Evangelos K. Oikonomou, Arya Aminorroaya, L. Dhingra, Caitlin Partridge, Eric J Velazquez, N. Desai, H. Krumholz, Edward J Miller, R. Khera
An algorithmic strategy for anatomical vs. functional testing in suspected coronary artery disease (CAD) (Anatomical vs. Stress teSting decIsion Support Tool; ASSIST) is associated with better outcomes than random selection. However, in the real world, this decision is rarely random. We explored the agreement between a provider-driven vs. simulated algorithmic approach to cardiac testing and its association with outcomes across multinational cohorts. In two cohorts of functional vs. anatomical testing in a US hospital health system [Yale; 2013–2023; n = 130 196 (97.0%) vs. n = 4020 (3.0%), respectively], and the UK Biobank [n = 3320 (85.1%) vs. n = 581 (14.9%), respectively], we examined outcomes stratified by agreement between the real-world and ASSIST-recommended strategies. Younger age, female sex, Black race, and diabetes history were independently associated with lower odds of ASSIST-aligned testing. Over a median of 4.9 (interquartile range [IQR]: 2.4–7.1) and 5.4 (IQR: 2.6–8.8) years, referral to the ASSIST-recommended strategy was associated with a lower risk of acute myocardial infarction or death (hazard ratioadjusted: 0.81, 95% confidence interval [CI] 0.77–0.85, P < 0.001 and 0.74 [95% CI 0.60–0.90], P = 0.003, respectively), an effect that remained significant across years, test types, and risk profiles. In post hoc analyses of anatomical-first testing in the Prospective Multicentre Imaging Study for Evaluation of Chest Pain (PROMISE) trial, alignment with ASSIST was independently associated with a 17% and 30% higher risk of detecting CAD in any vessel or the left main artery/proximal left anterior descending coronary artery, respectively. In cohorts where historical practices largely favour functional testing, alignment with an algorithmic approach to cardiac testing defined by ASSIST was associated with a lower risk of adverse outcomes. This highlights the potential utility of a data-driven approach in the diagnostic management of CAD.
对疑似冠状动脉疾病（CAD）进行解剖与功能测试的算法策略（解剖与压力测试决策支持工具；ASSIST）比随机选择的结果更好。然而，在现实世界中，这种决定很少是随机的。我们在多国队列中探讨了提供者驱动与模拟算法心脏测试方法之间的一致性及其与预后的关系。 在美国医院卫生系统[耶鲁大学；2013-2023 年；n = 130 196 (97.0%) vs. n = 4020 (3.0%)，]和英国生物库[n = 3320 (85.1%) vs. n = 581 (14.9%)，]的功能测试与解剖测试的两个队列中，我们根据真实世界与 ASSIST 推荐策略之间的一致性对结果进行了分层研究。年龄较小、女性、黑人和糖尿病史与较低的 ASSIST 一致测试几率独立相关。在中位数为 4.9（四分位间距 [IQR]：2.4-7.1）年和 5.4（IQR：2.6-8.8）年期间，转诊至 ASSIST 推荐策略与急性心肌梗死或死亡风险较低有关（调整后的危险比：0.81, 95% confidence interval [CI] 0.77-0.85, P < 0.001 and 0.74 [95% CI 0.60-0.90], P = 0.003, respectively)，这种效应在不同年份、检验类型和风险概况下都保持显著。在胸痛评估的前瞻性多中心成像研究（PROMISE）试验中，对解剖学优先检测进行的事后分析显示，与 ASSIST 保持一致与在任何血管或左主干/左前降支冠状动脉近端检测到 CAD 的风险分别高出 17% 和 30% 独立相关。 在历来偏重于功能检查的队列中，采用 ASSIST 所定义的心脏检查算法与较低的不良预后风险相关。这凸显了以数据为导向的方法在诊断管理 CAD 方面的潜在作用。
{"title":"Real-world evaluation of an algorithmic machine-learning-guided testing approach in stable chest pain: a multinational, multicohort study","authors":"Evangelos K. Oikonomou, Arya Aminorroaya, L. Dhingra, Caitlin Partridge, Eric J Velazquez, N. Desai, H. Krumholz, Edward J Miller, R. Khera","doi":"10.1093/ehjdh/ztae023","DOIUrl":"https://doi.org/10.1093/ehjdh/ztae023","url":null,"abstract":"\u0000 \u0000 \u0000 An algorithmic strategy for anatomical vs. functional testing in suspected coronary artery disease (CAD) (Anatomical vs. Stress teSting decIsion Support Tool; ASSIST) is associated with better outcomes than random selection. However, in the real world, this decision is rarely random. We explored the agreement between a provider-driven vs. simulated algorithmic approach to cardiac testing and its association with outcomes across multinational cohorts.\u0000 \u0000 \u0000 \u0000 In two cohorts of functional vs. anatomical testing in a US hospital health system [Yale; 2013–2023; n = 130 196 (97.0%) vs. n = 4020 (3.0%), respectively], and the UK Biobank [n = 3320 (85.1%) vs. n = 581 (14.9%), respectively], we examined outcomes stratified by agreement between the real-world and ASSIST-recommended strategies. Younger age, female sex, Black race, and diabetes history were independently associated with lower odds of ASSIST-aligned testing. Over a median of 4.9 (interquartile range [IQR]: 2.4–7.1) and 5.4 (IQR: 2.6–8.8) years, referral to the ASSIST-recommended strategy was associated with a lower risk of acute myocardial infarction or death (hazard ratioadjusted: 0.81, 95% confidence interval [CI] 0.77–0.85, P < 0.001 and 0.74 [95% CI 0.60–0.90], P = 0.003, respectively), an effect that remained significant across years, test types, and risk profiles. In post hoc analyses of anatomical-first testing in the Prospective Multicentre Imaging Study for Evaluation of Chest Pain (PROMISE) trial, alignment with ASSIST was independently associated with a 17% and 30% higher risk of detecting CAD in any vessel or the left main artery/proximal left anterior descending coronary artery, respectively.\u0000 \u0000 \u0000 \u0000 In cohorts where historical practices largely favour functional testing, alignment with an algorithmic approach to cardiac testing defined by ASSIST was associated with a lower risk of adverse outcomes. This highlights the potential utility of a data-driven approach in the diagnostic management of CAD.\u0000","PeriodicalId":508387,"journal":{"name":"European Heart Journal - Digital Health","volume":"113 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140731861","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Correction to: What drives performance in machine learning models for predicting heart failure outcome?","authors":"","doi":"10.1093/ehjdh/ztae019","DOIUrl":"https://doi.org/10.1093/ehjdh/ztae019","url":null,"abstract":"","PeriodicalId":508387,"journal":{"name":"European Heart Journal - Digital Health","volume":"98 44","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140377628","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Annemiek E van Ravensberg, Niels T B Scholte, Aaram Omar Khader, Jasper J Brugts, N. Bruining, Robert M A van der Boon
Early detection of congestion has demonstrated to improve outcomes in heart failure (HF) patients. However, there is limited-access to invasively hemodynamic parameters to guide treatment. This study aimed to develop a model to estimate the invasively measured pulmonary capillary wedge pressure (PCWP) using non-invasive measurements with both traditional statistics and machine learning (ML) techniques. The study involved patients undergoing right-sided heart catheterization at Erasmus MC, Rotterdam, from 2017 to 2022. Invasively measured PCWP served as outcomes. Model features included non-invasive measurements of arterial blood pressure, saturation, heart rate (variability), weight, and temperature. Various traditional and ML techniques were used, and performance was assessed with R² and AUC for regression and classification models, respectively. A total of 853 procedures were included of which 31% had HF as primary diagnosis and 49% had a PCWP of 12 mmHg or higher. The mean age of the cohort was 59 ± 14 years and 52% were male. The HRV had the highest correlation with the PCWP with a correlation of 0.16. All the regression models resulted in low R2 values of up to 0.04 and the classification models in AUC values of up to 0.59. In this study, non-invasive methods, both traditional and ML-based, showed limited correlation to PCWP. This highlights the weak correlation between traditional HF monitoring and hemodynamic parameters, also emphasizing the limitations of single non-invasive measurements. Future research should explore trend analysis and additional features to improve non-invasive hemodynamic monitoring, as there is a clear demand for further advancements in this field.
{"title":"Machine Learning-based Analysis of Non-Invasive Measurements for Predicting Intracardiac Pressures","authors":"Annemiek E van Ravensberg, Niels T B Scholte, Aaram Omar Khader, Jasper J Brugts, N. Bruining, Robert M A van der Boon","doi":"10.1093/ehjdh/ztae021","DOIUrl":"https://doi.org/10.1093/ehjdh/ztae021","url":null,"abstract":"\u0000 \u0000 \u0000 Early detection of congestion has demonstrated to improve outcomes in heart failure (HF) patients. However, there is limited-access to invasively hemodynamic parameters to guide treatment. This study aimed to develop a model to estimate the invasively measured pulmonary capillary wedge pressure (PCWP) using non-invasive measurements with both traditional statistics and machine learning (ML) techniques.\u0000 \u0000 \u0000 \u0000 The study involved patients undergoing right-sided heart catheterization at Erasmus MC, Rotterdam, from 2017 to 2022. Invasively measured PCWP served as outcomes. Model features included non-invasive measurements of arterial blood pressure, saturation, heart rate (variability), weight, and temperature. Various traditional and ML techniques were used, and performance was assessed with R² and AUC for regression and classification models, respectively.\u0000 \u0000 \u0000 \u0000 A total of 853 procedures were included of which 31% had HF as primary diagnosis and 49% had a PCWP of 12 mmHg or higher. The mean age of the cohort was 59 ± 14 years and 52% were male. The HRV had the highest correlation with the PCWP with a correlation of 0.16. All the regression models resulted in low R2 values of up to 0.04 and the classification models in AUC values of up to 0.59.\u0000 \u0000 \u0000 \u0000 In this study, non-invasive methods, both traditional and ML-based, showed limited correlation to PCWP. This highlights the weak correlation between traditional HF monitoring and hemodynamic parameters, also emphasizing the limitations of single non-invasive measurements. Future research should explore trend analysis and additional features to improve non-invasive hemodynamic monitoring, as there is a clear demand for further advancements in this field.\u0000","PeriodicalId":508387,"journal":{"name":"European Heart Journal - Digital Health","volume":"21 4 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140245195","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Milagros Pereyra, Juan Farina, Ahmed K. Mahmoud, Isabel G. Scalia, Francesca Galasso, Michael E Killian, Mustafa Suppah, Courtney R Kenyon, Laura M Koepke, R. Padang, Chieh-Ju Chao, John P Sweeney, F. Fortuin, M. Eleid, Kristen A. Sell-Dottin, D. Steidley, Luis R. Scott, Rafael Fonseca, Francisco Lopez-Jimenez, Z. Attia, A. Dispenzieri, M. Grogan, Julie L. Rosenthal, R. Arsanjani, Chadi Ayoub
Cardiac amyloidosis (CA) is common in patients with severe aortic stenosis (AS) undergoing transcatheter aortic valve replacement (TAVR). CA has poor outcomes, and its assessment in all TAVR patients is costly and challenging. Electrocardiogram (ECG) artificial intelligence (AI) algorithms that screen for CA may be useful to identify at risk patients. In this retrospective analysis of our institutional National Cardiovascular Disease Registry (NCDR)-TAVR database, patients undergoing TAVR between January 2012 and December 2018 were included. Pre-TAVR CA probability was analyzed by an ECG AI predictive model, with >50% risk defined as high probability for CA. Univariable and propensity score covariate adjustment analysis using Cox regression was performed to compare clinical outcomes between patients with high CA probability versus those with low probability at one year follow-up after TAVR. Of 1426 patients who underwent TAVR (mean age 81.0 ± 8.5 years, 57.6% male), 349 (24.4%) had high CA probability on pre-procedure ECG. Only 17 (1.2%) had clinical diagnosis of CA. After multivariable adjustment, high probability of CA by ECG-AI algorithm was significantly associated with increased all-cause mortality (HR 1.40, 95%CI 1.01-1.96, p = 0.046) and higher rates of MACE (TIA/Stroke, myocardial infarction, heart failure hospitalizations) (HR 1.36, 95%CI 1.01- 1.82, p = 0.041), driven primarily by heart failure hospitalizations (HR 1.58, 95%CI 1.13-2.20, p = 0.008) at one-year follow-up. There were no significant differences in TIA/Stroke or myocardial infarction. AI applied to pre-TAVR ECGs identifies a subgroup at higher risk of clinical events. These targeted patients may benefit from further diagnostic evaluation for CA.
{"title":"The Prognostic Value of Artificial Intelligence to Predict Cardiac Amyloidosis in Patients with Severe Aortic Stenosis Undergoing Transcatheter Aortic Valve Replacement","authors":"Milagros Pereyra, Juan Farina, Ahmed K. Mahmoud, Isabel G. Scalia, Francesca Galasso, Michael E Killian, Mustafa Suppah, Courtney R Kenyon, Laura M Koepke, R. Padang, Chieh-Ju Chao, John P Sweeney, F. Fortuin, M. Eleid, Kristen A. Sell-Dottin, D. Steidley, Luis R. Scott, Rafael Fonseca, Francisco Lopez-Jimenez, Z. Attia, A. Dispenzieri, M. Grogan, Julie L. Rosenthal, R. Arsanjani, Chadi Ayoub","doi":"10.1093/ehjdh/ztae022","DOIUrl":"https://doi.org/10.1093/ehjdh/ztae022","url":null,"abstract":"\u0000 \u0000 \u0000 Cardiac amyloidosis (CA) is common in patients with severe aortic stenosis (AS) undergoing transcatheter aortic valve replacement (TAVR). CA has poor outcomes, and its assessment in all TAVR patients is costly and challenging. Electrocardiogram (ECG) artificial intelligence (AI) algorithms that screen for CA may be useful to identify at risk patients.\u0000 \u0000 \u0000 \u0000 In this retrospective analysis of our institutional National Cardiovascular Disease Registry (NCDR)-TAVR database, patients undergoing TAVR between January 2012 and December 2018 were included. Pre-TAVR CA probability was analyzed by an ECG AI predictive model, with >50% risk defined as high probability for CA. Univariable and propensity score covariate adjustment analysis using Cox regression was performed to compare clinical outcomes between patients with high CA probability versus those with low probability at one year follow-up after TAVR.\u0000 \u0000 \u0000 \u0000 Of 1426 patients who underwent TAVR (mean age 81.0 ± 8.5 years, 57.6% male), 349 (24.4%) had high CA probability on pre-procedure ECG. Only 17 (1.2%) had clinical diagnosis of CA. After multivariable adjustment, high probability of CA by ECG-AI algorithm was significantly associated with increased all-cause mortality (HR 1.40, 95%CI 1.01-1.96, p = 0.046) and higher rates of MACE (TIA/Stroke, myocardial infarction, heart failure hospitalizations) (HR 1.36, 95%CI 1.01- 1.82, p = 0.041), driven primarily by heart failure hospitalizations (HR 1.58, 95%CI 1.13-2.20, p = 0.008) at one-year follow-up. There were no significant differences in TIA/Stroke or myocardial infarction.\u0000 \u0000 \u0000 \u0000 AI applied to pre-TAVR ECGs identifies a subgroup at higher risk of clinical events. These targeted patients may benefit from further diagnostic evaluation for CA.\u0000","PeriodicalId":508387,"journal":{"name":"European Heart Journal - Digital Health","volume":"2020 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140246017","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Niels T B Scholte, Annemiek E van Ravensberg, Roos Edgar, A. J. van den Enden, Nicolas van Mieghem, Jasper J Brugts, J. Bonnes, N. Bruining, R. M. van der Boon
Invasive hemodynamic monitoring of heart failure (HF) is used to detect deterioration in an early phase thereby preventing hospitalizations. However, this invasive approach is costly and presently lacks widespread accessibility. Hence, there is a pressing need to identify an alternative non-invasive method that is reliable and more readily available. In this pilot study we investigated the relation between wrist-derived Photoplethysmography (PPG) signals and the invasively measured pulmonary capillary wedge pressure (PCWP). Fourteen patients with aortic valve stenosis who underwent Transcatheter Aortic Valve Replacement with concomitant right heart catheterization and PPG measurements were included. Six unique features of the PPG signals (heart rate, heart rate variability, systolic amplitude (SA), diastolic amplitude, crest time (CT), and large artery stiffness index (LASI) were extracted. These features were used to estimate the continuous PCWP values and the categorized PCWP (low <12mmHg vs. high ≥12mmHg). All PPG features resulted in regression models that showed low correlations with the invasively measured PCWP. Classification models resulted in higher performances: the model based on the SA and the model based on the LASI both resulted in an Area Under the Curve(AUC) of 0.86 and the model based on the CT resulted in an AUC of 0.72. These results demonstrate the capability to non-invasively classify patients into clinically meaningful categories of PCWP using PPG signals from a wrist-worn wearable device. To enhance and fully explore its potential, the relationship between PPG and PCWP should be further investigated in a larger cohort of HF patients.
{"title":"Photoplethysmography and Intracardiac Pressures: early insights from a pilot study","authors":"Niels T B Scholte, Annemiek E van Ravensberg, Roos Edgar, A. J. van den Enden, Nicolas van Mieghem, Jasper J Brugts, J. Bonnes, N. Bruining, R. M. van der Boon","doi":"10.1093/ehjdh/ztae020","DOIUrl":"https://doi.org/10.1093/ehjdh/ztae020","url":null,"abstract":"\u0000 \u0000 \u0000 Invasive hemodynamic monitoring of heart failure (HF) is used to detect deterioration in an early phase thereby preventing hospitalizations. However, this invasive approach is costly and presently lacks widespread accessibility. Hence, there is a pressing need to identify an alternative non-invasive method that is reliable and more readily available. In this pilot study we investigated the relation between wrist-derived Photoplethysmography (PPG) signals and the invasively measured pulmonary capillary wedge pressure (PCWP).\u0000 \u0000 \u0000 \u0000 Fourteen patients with aortic valve stenosis who underwent Transcatheter Aortic Valve Replacement with concomitant right heart catheterization and PPG measurements were included. Six unique features of the PPG signals (heart rate, heart rate variability, systolic amplitude (SA), diastolic amplitude, crest time (CT), and large artery stiffness index (LASI) were extracted. These features were used to estimate the continuous PCWP values and the categorized PCWP (low <12mmHg vs. high ≥12mmHg). All PPG features resulted in regression models that showed low correlations with the invasively measured PCWP. Classification models resulted in higher performances: the model based on the SA and the model based on the LASI both resulted in an Area Under the Curve(AUC) of 0.86 and the model based on the CT resulted in an AUC of 0.72.\u0000 \u0000 \u0000 \u0000 These results demonstrate the capability to non-invasively classify patients into clinically meaningful categories of PCWP using PPG signals from a wrist-worn wearable device. To enhance and fully explore its potential, the relationship between PPG and PCWP should be further investigated in a larger cohort of HF patients.\u0000","PeriodicalId":508387,"journal":{"name":"European Heart Journal - Digital Health","volume":"46 6","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-03-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140259222","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
S. Papadopoulou, D. Dionysopoulos, V. Mentesidou, K. Loga, S. Michalopoulou, C. Koukoutzeli, K. Efthimiadis, V. Kantartzi, E. Timotheadou, I. Styliadis, P. Nihoyannopoulos, V. Sachpekidis
Left ventricular ejection fraction (LVEF) calculation by echocardiography is pivotal in evaluating cancer patients’ cardiac function. Artificial intelligence (AI) can facilitate acquisition of optimal images and automated LVEF calculation. We sought to evaluate the feasibility and accuracy of LVEF calculation by oncology staff using an AI-enabled handheld ultrasound device (HUD). We studied 115 patients referred for echocardiographic LVEF estimation. All patients were scanned by a cardiologist using standard echocardiography (SE) and biplane Simpson’s LVEF was the reference standard. Hands-on training using the Kosmos HUD was provided to the oncology staff before the study. Each patient was scanned by a cardiologist, a senior oncologist, an oncology resident, and a nurse using the TRIO AI and KOSMOS EF deep learning algorithms to obtain automated LVEF (autoEF). The correlation between autoEF and SE-EF was excellent for the cardiologist (r = 0.90), the junior oncologist (r = 0.82) and the nurse (r = 0.84), and good for the senior oncologist (r = 0.79). The Bland-Altman analysis showed small underestimation by autoEF compared to SE-EF. Detection of impaired LVEF < 50% was feasible with sensitivity 95% and specificity 94% for the cardiologist; sensitivity 86% and specificity 93% for the senior oncologist; sensitivity 95% and specificity 91% for the junior oncologist; sensitivity 94% and specificity 87% for the nurse. Automated LVEF calculation by oncology staff was feasible using AI-enabled HUD in a selected patient population. Detection of LVEF < 50% was possible with good accuracy. These findings show potential to expedite clinical workflow of cancer patients and speed up referral when necessary.
{"title":"Artificial Intelligence-assisted evaluation of cardiac function by oncology staff in chemotherapy patients","authors":"S. Papadopoulou, D. Dionysopoulos, V. Mentesidou, K. Loga, S. Michalopoulou, C. Koukoutzeli, K. Efthimiadis, V. Kantartzi, E. Timotheadou, I. Styliadis, P. Nihoyannopoulos, V. Sachpekidis","doi":"10.1093/ehjdh/ztae017","DOIUrl":"https://doi.org/10.1093/ehjdh/ztae017","url":null,"abstract":"\u0000 \u0000 \u0000 Left ventricular ejection fraction (LVEF) calculation by echocardiography is pivotal in evaluating cancer patients’ cardiac function. Artificial intelligence (AI) can facilitate acquisition of optimal images and automated LVEF calculation. We sought to evaluate the feasibility and accuracy of LVEF calculation by oncology staff using an AI-enabled handheld ultrasound device (HUD).\u0000 \u0000 \u0000 \u0000 We studied 115 patients referred for echocardiographic LVEF estimation. All patients were scanned by a cardiologist using standard echocardiography (SE) and biplane Simpson’s LVEF was the reference standard. Hands-on training using the Kosmos HUD was provided to the oncology staff before the study. Each patient was scanned by a cardiologist, a senior oncologist, an oncology resident, and a nurse using the TRIO AI and KOSMOS EF deep learning algorithms to obtain automated LVEF (autoEF).\u0000 \u0000 \u0000 \u0000 The correlation between autoEF and SE-EF was excellent for the cardiologist (r = 0.90), the junior oncologist (r = 0.82) and the nurse (r = 0.84), and good for the senior oncologist (r = 0.79). The Bland-Altman analysis showed small underestimation by autoEF compared to SE-EF. Detection of impaired LVEF < 50% was feasible with sensitivity 95% and specificity 94% for the cardiologist; sensitivity 86% and specificity 93% for the senior oncologist; sensitivity 95% and specificity 91% for the junior oncologist; sensitivity 94% and specificity 87% for the nurse.\u0000 \u0000 \u0000 \u0000 Automated LVEF calculation by oncology staff was feasible using AI-enabled HUD in a selected patient population. Detection of LVEF < 50% was possible with good accuracy. These findings show potential to expedite clinical workflow of cancer patients and speed up referral when necessary.\u0000","PeriodicalId":508387,"journal":{"name":"European Heart Journal - Digital Health","volume":"12 3","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140425875","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
J. Naser, E. Lee, S. Pislaru, Gal Tsaban, Jeffrey G Malins, John I Jackson, D. Anisuzzaman, Behrouz Rostami, Francisco Lopez-Jimenez, Paul A. Friedman, Garvan C. Kane, Patricia A. Pellikka, Z. Attia
Augmenting echocardiography with artificial intelligence would allow for automated assessment of routine parameters and identification of disease patterns not easily recognized otherwise. View classification is an essential first step before deep learning can be applied to the echocardiogram. We trained 2- and 3-dimensional convolutional neural networks (CNNs) using transthoracic echocardiographic (TTE) studies obtained from 909 patients to classify 9 view categories [10,269 videos]. TTE studies from 229 patients were used in internal validation [2,582 videos]. CNNs were tested on 100 patients with comprehensive TTE studies [where the 2 examples chosen by CNNs as most likely to represent a view were evaluated] and 408 patients with five view categories obtained via point of care ultrasound (POCUS). The overall accuracy of the 2-dimensional CNN was 96.8% and the averaged area under the curve (AUC) was 0.997 on the comprehensive TTE testing set; these numbers were 98.4% and 0.998, respectively, on the POCUS set. For the 3-dimensional CNN, the accuracy and AUC were 96.3% and 0.998 for full TTE studies and 95.0% and 0.996 on POCUS videos, respectively. The positive predictive value, which defined correctly identified predicted views, was higher with 2- rather than 3-dimensional networks, exceeding 93% in apical, short axis aortic valve, and parasternal long axis left ventricle views. An automated view classifier utilizing CNNs was able to classify cardiac views obtained using TTE and POCUS with high accuracy. The view classifier will facilitate the application of deep learning to echocardiography.
{"title":"Artificial intelligence-based classification of echocardiographic views","authors":"J. Naser, E. Lee, S. Pislaru, Gal Tsaban, Jeffrey G Malins, John I Jackson, D. Anisuzzaman, Behrouz Rostami, Francisco Lopez-Jimenez, Paul A. Friedman, Garvan C. Kane, Patricia A. Pellikka, Z. Attia","doi":"10.1093/ehjdh/ztae015","DOIUrl":"https://doi.org/10.1093/ehjdh/ztae015","url":null,"abstract":"\u0000 \u0000 \u0000 Augmenting echocardiography with artificial intelligence would allow for automated assessment of routine parameters and identification of disease patterns not easily recognized otherwise. View classification is an essential first step before deep learning can be applied to the echocardiogram.\u0000 \u0000 \u0000 \u0000 We trained 2- and 3-dimensional convolutional neural networks (CNNs) using transthoracic echocardiographic (TTE) studies obtained from 909 patients to classify 9 view categories [10,269 videos]. TTE studies from 229 patients were used in internal validation [2,582 videos]. CNNs were tested on 100 patients with comprehensive TTE studies [where the 2 examples chosen by CNNs as most likely to represent a view were evaluated] and 408 patients with five view categories obtained via point of care ultrasound (POCUS).\u0000 \u0000 \u0000 \u0000 The overall accuracy of the 2-dimensional CNN was 96.8% and the averaged area under the curve (AUC) was 0.997 on the comprehensive TTE testing set; these numbers were 98.4% and 0.998, respectively, on the POCUS set. For the 3-dimensional CNN, the accuracy and AUC were 96.3% and 0.998 for full TTE studies and 95.0% and 0.996 on POCUS videos, respectively. The positive predictive value, which defined correctly identified predicted views, was higher with 2- rather than 3-dimensional networks, exceeding 93% in apical, short axis aortic valve, and parasternal long axis left ventricle views.\u0000 \u0000 \u0000 \u0000 An automated view classifier utilizing CNNs was able to classify cardiac views obtained using TTE and POCUS with high accuracy. The view classifier will facilitate the application of deep learning to echocardiography.\u0000","PeriodicalId":508387,"journal":{"name":"European Heart Journal - Digital Health","volume":"16 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140429917","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
P. Sultanian, P. Lundgren, Antros Louca, Erik Andersson, Therese Djärv, Fredrik Hessulf, Anna Henningsson, A. Martinsson, P. Nordberg, Adam Piasecki, Vibha Gupta, Z. Mandalenakis, Amar Taha, Bengt Redfors, Johan Herlitz, A. Rawshani
Out-of-hospital cardiac arrest (OHCA) is a major health concern worldwide. Although one third of all cases achieve return of spontaneous circulation (ROSC) and may undergo a difficult period in the ICU, only one in ten survive. This study aimed to improve our previously developed machine learning model for early prognostication of survival in OHCA. We studied all cases registered in the Swedish Cardiopulmonary Resuscitation Registry during 2010 and 2020 (n=55,615). We compared the predictive performance of extreme gradient boosting (XGB), LightGBM, logistic regression, CatBoost, random forest and TabNet. For each framework, we developed models that optimized (1) a weighted F1 score to penalize models that yielded more false negatives, and (2) PR AUC (precision recall area under the curve). LightGBM assigned higher importance values to a larger set of variables, while XGB made predictions using fewer predictors. The AUC ROC scores for LightGBM was 0.958 (optimized for weighted F1) and 0.961 (optimized for PR AUC), while for XGB, the scores were 0.958 and 0.960 respectively. The calibration plots showed subtle underestimation of survival for LightGBM, contrasting with a mild overestimation for XGB models. In the crucial range of 0 to 10% likelihood of survival, the XGB model, optimized with PR AUC, emerged as a clinically safe model. We improved our previous prediction model by creating a parsimonious model with AUC ROC at 0.96, with excellent calibration and no apparent risk of underestimating survival in the critical probability range (0-10%). The model is available at www.gocares.se.
{"title":"Prediction of survival in out-of-hospital cardiac arrest: The updated SCARS Model","authors":"P. Sultanian, P. Lundgren, Antros Louca, Erik Andersson, Therese Djärv, Fredrik Hessulf, Anna Henningsson, A. Martinsson, P. Nordberg, Adam Piasecki, Vibha Gupta, Z. Mandalenakis, Amar Taha, Bengt Redfors, Johan Herlitz, A. Rawshani","doi":"10.1093/ehjdh/ztae016","DOIUrl":"https://doi.org/10.1093/ehjdh/ztae016","url":null,"abstract":"\u0000 \u0000 \u0000 Out-of-hospital cardiac arrest (OHCA) is a major health concern worldwide. Although one third of all cases achieve return of spontaneous circulation (ROSC) and may undergo a difficult period in the ICU, only one in ten survive. This study aimed to improve our previously developed machine learning model for early prognostication of survival in OHCA.\u0000 \u0000 \u0000 \u0000 We studied all cases registered in the Swedish Cardiopulmonary Resuscitation Registry during 2010 and 2020 (n=55,615). We compared the predictive performance of extreme gradient boosting (XGB), LightGBM, logistic regression, CatBoost, random forest and TabNet. For each framework, we developed models that optimized (1) a weighted F1 score to penalize models that yielded more false negatives, and (2) PR AUC (precision recall area under the curve).\u0000 \u0000 \u0000 \u0000 LightGBM assigned higher importance values to a larger set of variables, while XGB made predictions using fewer predictors. The AUC ROC scores for LightGBM was 0.958 (optimized for weighted F1) and 0.961 (optimized for PR AUC), while for XGB, the scores were 0.958 and 0.960 respectively. The calibration plots showed subtle underestimation of survival for LightGBM, contrasting with a mild overestimation for XGB models. In the crucial range of 0 to 10% likelihood of survival, the XGB model, optimized with PR AUC, emerged as a clinically safe model.\u0000 \u0000 \u0000 \u0000 We improved our previous prediction model by creating a parsimonious model with AUC ROC at 0.96, with excellent calibration and no apparent risk of underestimating survival in the critical probability range (0-10%). The model is available at www.gocares.se.\u0000","PeriodicalId":508387,"journal":{"name":"European Heart Journal - Digital Health","volume":"40 5","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140434893","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Juan Lu, A. Bisson, Mohammed Bennamoun, Yalin Zheng, Frank M. Sanfilippo, Joseph Hung, Tom Briffa, Brendan M. McQuillan, J. Stewart, Gemma Figtree, M. V. Huisman, Girish Dwivedi, G. Y. Lip
Patients with atrial fibrillation (AF) have a higher risk of ischaemic stroke and death. While anticoagulants are effective at reducing these risks, they increase the risk of bleeding. Current clinical risk scores only perform modestly in predicting adverse outcomes, especially for the outcome of death. We aimed to test the multi-label gradient boosting decision tree (ML-GBDT) model in predicting risks for adverse outcomes in a prospective global AF registry. We studied patients from phase II/III of the GLORIA-AF registry between 2011 and 2020. The outcomes were all-cause death, ischaemic stroke and major bleeding within one year following the AF. We trained the ML-GBDT model and compared its discrimination with the clinical scores in predicting patient outcomes. A total of 25,656 patients were included (mean age 70.3 years (SD 10.3); 44.8% female). Within one-year after AF, ischaemic stroke occurred in 215 (0.8%), major bleeding in 405 (1.6%), and death in 897 (3.5%) patients. Our model achieved an optimized area under the curve (AUC) in predicting death (0.785, 95% CI: 0.757-0.813) compared to Charlson Comorbidity Index (0.747, p=0.007), ischaemic stroke (0.691, 0.626-0.756) comparing to CHA2DS2-VASc (0.613, p=0.028), and major bleeding (0.698, 0.651-0.745) as opposed to HAS-BLED (0.607, p=0.002), with improvement in net reclassification index (10.0%, 12.5% and 23.6% respectively). The ML-GBDT model outperformed clinical risk scores in predicting the risks in patients with AF. This approach could be used as a single multifaceted holistic tool to optimize patient risk assessment and mitigate adverse outcomes when managing AF.
{"title":"Predicting multifaceted risks using machine learning in Atrial Fibrillation: Insights from GLORIA-AF study","authors":"Juan Lu, A. Bisson, Mohammed Bennamoun, Yalin Zheng, Frank M. Sanfilippo, Joseph Hung, Tom Briffa, Brendan M. McQuillan, J. Stewart, Gemma Figtree, M. V. Huisman, Girish Dwivedi, G. Y. Lip","doi":"10.1093/ehjdh/ztae010","DOIUrl":"https://doi.org/10.1093/ehjdh/ztae010","url":null,"abstract":"\u0000 \u0000 \u0000 Patients with atrial fibrillation (AF) have a higher risk of ischaemic stroke and death. While anticoagulants are effective at reducing these risks, they increase the risk of bleeding. Current clinical risk scores only perform modestly in predicting adverse outcomes, especially for the outcome of death. We aimed to test the multi-label gradient boosting decision tree (ML-GBDT) model in predicting risks for adverse outcomes in a prospective global AF registry.\u0000 \u0000 \u0000 \u0000 We studied patients from phase II/III of the GLORIA-AF registry between 2011 and 2020. The outcomes were all-cause death, ischaemic stroke and major bleeding within one year following the AF. We trained the ML-GBDT model and compared its discrimination with the clinical scores in predicting patient outcomes. A total of 25,656 patients were included (mean age 70.3 years (SD 10.3); 44.8% female). Within one-year after AF, ischaemic stroke occurred in 215 (0.8%), major bleeding in 405 (1.6%), and death in 897 (3.5%) patients. Our model achieved an optimized area under the curve (AUC) in predicting death (0.785, 95% CI: 0.757-0.813) compared to Charlson Comorbidity Index (0.747, p=0.007), ischaemic stroke (0.691, 0.626-0.756) comparing to CHA2DS2-VASc (0.613, p=0.028), and major bleeding (0.698, 0.651-0.745) as opposed to HAS-BLED (0.607, p=0.002), with improvement in net reclassification index (10.0%, 12.5% and 23.6% respectively).\u0000 \u0000 \u0000 \u0000 The ML-GBDT model outperformed clinical risk scores in predicting the risks in patients with AF. This approach could be used as a single multifaceted holistic tool to optimize patient risk assessment and mitigate adverse outcomes when managing AF.\u0000","PeriodicalId":508387,"journal":{"name":"European Heart Journal - Digital Health","volume":"289 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140448133","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：