scholarly journals A Phenotyping of Diastolic Function by Machine Learning Improves Prediction of Clinical Outcomes in Heart Failure

2021 ◽  
Vol 8 ◽  
Author(s):  
Haruka Kameshima ◽  
Tokuhisa Uejima ◽  
Alan G. Fraser ◽  
Lisa Takahashi ◽  
Junyi Cho ◽  
...  
Keyword(s):  
Machine Learning ◽  
Heart Failure ◽  
Diastolic Function ◽  
Clinical Outcomes ◽  
Primary Endpoint ◽  
Learning Algorithm ◽  
Global Longitudinal Strain ◽  
Left Ventricular ◽  
Hazard Ratio ◽  
Homogeneous Groups

Background: Discriminating between different patterns of diastolic dysfunction in heart failure (HF) is still challenging. We tested the hypothesis that an unsupervised machine learning algorithm would detect heterogeneity in diastolic function and improve risk stratification compared with recommended consensus criteria.Methods: This study included 279 consecutive patients aged 24–97 years old with clinically stable HF referred for echocardiographic assessment, in whom diastolic variables were measured according to the current guidelines. Cluster analysis was undertaken to identify homogeneous groups of patients with similar profiles of the variables. Sequential Cox models were used to compare cluster-based classification with guidelines-based classification for predicting clinical outcomes. The primary endpoint was hospitalization for worsening HF.Results: The analysis identified three clusters with distinct properties of diastolic function that shared similarities with guidelines-based classification. The clusters were associated with brain natriuretic peptide level (p < 0.001), hemoglobin concentration (p = 0.017) and estimated glomerular filtration rate (p = 0.001). During a mean follow-up period of 2.6 ± 2.0 years, 62 patients (22%) experienced the primary endpoint. Cluster-based classification predicted events with a hazard ratio 1.68 (p = 0.019) that was independent from and incremental to the Meta-analysis Global Group in Chronic Heart Failure (MAGGIC) risk score for HF, and from left ventricular end-diastolic volume and global longitudinal strain, whereas guidelines-based classification did not retain its independent prognostic value (hazard ratio = 1.25, p = 0.202).Conclusion: Machine learning can identify patterns of diastolic function that better stratify the risk for decompensation than the current consensus recommendations in HF. Integrating this data-driven phenotyping may help in refining prognostication and optimizing treatment.

scholarly journals 153 Multiparametric assessment of diastolic function in heart failure

2020 ◽  
Vol 21 (Supplement_1) ◽  
Author(s):  
T Uejima ◽  
J Cho ◽  
H Hayama ◽  
L Takahashi ◽  
J Yajima ◽  
...  
Keyword(s):  
Machine Learning ◽  
Heart Failure ◽  
Cluster Analysis ◽  
Diastolic Function ◽  
Primary Endpoint ◽  
Prognostic Value ◽  
Learning Algorithm ◽  
Left Ventricular ◽  
Machine Learning Techniques ◽  
Machine Learning Algorithm

Abstract Background The assessment of diastolic function is still challenging in the setting of heart failure (HF). We tested the hypothesis that applying a machine learning algorithm would detect heterogeneity in diastolic function and improve risk stratification in HF population. Methods This study included consecutive 279 patients with clinically stable HF referred for echocardiographic assessment, for whom diastolic function variables were measured according to the current guidelines. Cluster analysis, an unsupervised machine learning algorithm, was undertaken on these variables to form homogeneous groups of patients with similar profiles of the variables. Sequential Cox models paralleling the clinical sequence of HF assessment were used to elucidate the benefit of cluster-based classification over guidelines-based classification. The primary endpoint was a hospitalization for worsening HF. Results Cluster analysis identified 3 clusters with distinct properties of diastolic function that shared similarities with guidelines-based classification. The clusters were associated with brain natriuretic peptide level (p < 0.001, figure A). During follow-up period of 2.6 ± 2.0 years, 62 patients (22%) experienced the primary endpoint. Cluster-based classification exhibited a significant prognostic value (c2 = 20.3, p < 0.001, figure B), independent from and incremental to an established clinical risk score for HF (MAGGIC score) and left ventricular end-diastolic volume (hazard ratio = 1.677, p = 0.017, model c2: from 47.5 to 54.1, p = 0.015, figure D). Although guideline-based classification showed a significant prognostic value (c2 = 13.1, p = 0.001, figure C), it did not significantly improve overall prognostication from the baseline (model c2: from 47.5 to 49.9, p = 0.199, figure D). Conclusion Machine learning techniques help grading diastolic function and stratifying the risk for decompensation in HF. Abstract 153 Figure.

scholarly journals P893 Left ventricular flow energetics predicts worsening heart failure in dilated cardiomyopathy

2020 ◽  
Vol 21 (Supplement_1) ◽  
Author(s):  
L Takahashi ◽  
T Uejima ◽  
H Hayama ◽  
J Cho ◽  
T Chikamori ◽  
...  
Keyword(s):  
Heart Failure ◽  
Dilated Cardiomyopathy ◽  
Diastolic Function ◽  
Clinical Outcomes ◽  
Left Ventricular ◽  
Hazard Ratio ◽  
Total Flow ◽  
Blood Flows ◽  
Kaplan Meier ◽  
Significant Difference

Abstract Background Blood flows through healthy hearts form optimal flow structures; they store flow kinetic energy (KE) that can be used for ejection. In contrast, in failing hearts, intracardiac flows become disorganized so that they may be energetically inefficient. However, it remained unknown whether left ventricular (LV) flow energetics prognosticates in heart failure. Methods This study included 61 patients with dilated cardiomyopathy (DCM). The temporal change in KE during early diastole (ED), atrial contraction (AC) and isovolumic contraction (IVC) was measured using Vector Flow Mapping particle tracking (Hitachi, figure top). LV inflow (total flow) were divided, based on whether they were ejected (direct flow, DF) or stayed in LV (retained flow, RF) in the following systole. KE of DF can be made use of for ejection, whereas KE of RF is supposed to be wasted. Diastolic function was graded, according to current EACVI/ASE guidelines. The patients were followed up for three years. Primary endpoint was hospitalization for worsening heart failure (WHF). Results 12 patients had hospitalizations for WHF in the follow-up period. KE of total flow did not show any significant difference through the cardiac cycle between patients with and without WHF. KE of DF was slightly, but not significantly, smaller (ED: p = 0.252, AC: p = 0.119, IVC: p = 0.122), and KE of RF was slightly, but not significantly, larger (ED: p = 0.971, AC: p = 0.085, IVC: p = 0.134) in patients with WHF than those without events. The ratio of DF and RF (DF/RF ratio) showed significant differences between these two groups, especially from AC through IVC (figure, bottom-left). Cox proportional hazard analyses demonstrated that DF/RF ratio during IVC showed a significant correlation with clinical outcomes (p = 0.033, hazard ratio = 0.067). It remained significant even after adjusted for diastolic function grade (p = 0.046, hazard ratio = 0.074). Kaplan-Meier analysis confirmed the above results (figure, bottom-right). Conclusion: Efficiency of KE recruitment for LV ejection during IVC is associated with clinical outcomes in DCM. Abstract P893 Figure. LISA

scholarly journals Application of the current HFA-ESC consensus guidelines on diagnosis of heart failure with preserved ejection fraction to a population referred for echocardiography

2021 ◽  
Vol 22 (Supplement_1) ◽  
Author(s):  
K Liang ◽  
R Hearse-Morgan ◽  
S Fairbairn ◽  
Y Ismail ◽  
AK Nightingale
Keyword(s):  
Heart Failure ◽  
Atrial Fibrillation ◽  
Ejection Fraction ◽  
Diastolic Function ◽  
Global Longitudinal Strain ◽  
Diagnostic Algorithm ◽  
Left Ventricular ◽  
Lv Function ◽  
Preserved Ejection Fraction ◽  
Consensus Guidelines

Abstract Funding Acknowledgements Type of funding sources: None. BACKGROUND The recent Heart Failure Association (HFA) of the European Society of Cardiology (ESC) consensus guidelines on diagnosis of heart failure with preserved ejection fraction (HFpEF) have developed a simple diagnostic algorithm for clinical use. PURPOSE To assess whether echocardiogram (echo) parameters needed to assess diastolic function are routinely collected in patients referred for assessment of heart failure symptoms. METHODS Retrospective analysis of echo referrals in January 2020 were assessed for parameters of diastolic function as per step 2 of the HF-PEFF diagnostic algorithm.  Echo images and clinical reports were reviewed. Electronic records were utilised to obtain clinical history, blood results (NT-proBNP) and demographic data. RESULTS 1330 patients underwent an echo in our department during January 2020. 83 patients were referred with symptoms of heart failure without prior history of cardiac disease; 20 patients found to have impaired left ventricular (LV) function were excluded from analysis. Of the 63 patients with possible HFpEF, HF-PEFF score was low in 18, intermediate in 33 and high in 12. Median age was 68 years (range 32 to 97 years); 25% had a BMI >30. There was a high prevalence of hypertension (52%), diabetes (19%) and atrial fibrillation (40%) (cf. Table 1). Body surface area (BSA) was documented in 65% of echo reports. Most echo parameters were recorded with the exception of global longitudinal strain (GLS) and indexed LV mass (cf. image 1). NT-proBNP was recorded in only 20 patients (31.7%). 12 patients with an intermediate HF-PEFF score could have been re-categorised to a high score depending on GLS and NT-proBNP (which were not recorded). CONCLUSION More than three quarters of echoes acquired in our department obtained the relevant parameters to assess diastolic function. The addition of BSA, and inclusion of NT-proBNP, and GLS would have been additive to a third of ‘intermediate’ patients to determine definite HFpEF. Our study demonstrates that the current HFA-ESC diagnostic algorithm and HF-PEFF scoring system are easy to use, highly relevant and applicable to current clinical practice. Age >70 years 29 (46.0%) Obesity (BMI >30) 16 (25.4%) Diabetes 12 (19%) Hypertension 33 (52.4%) Atrial Fibrillation 25 (39.7%) ECG abnormalities 18 (28.5%) Table 1. Prevalence of Clinical Risk Factors Abstract Figure. Image 1. HFPEFF score & echo parameters

Multiomics, virtual reality and artificial intelligence in heart failure

2021 ◽  
Author(s):  
Patrick A Gladding ◽  
Suzanne Loader ◽  
Kevin Smith ◽  
Erica Zarate ◽  
Saras Green ◽  
...  
Keyword(s):  
Artificial Intelligence ◽  
Machine Learning ◽  
Heart Failure ◽  
Virtual Reality ◽  
Mental Stress ◽  
Longitudinal Strain ◽  
Stress Test ◽  
Global Longitudinal Strain ◽  
Area Under The Curve ◽  
Left Ventricular

Aim: Multiomics delivers more biological insight than targeted investigations. We applied multiomics to patients with heart failure (HF) and reduced ejection fraction (HFrEF), with machine learning applied to advanced ECG (AECG) and echocardiography artificial intelligence (Echo AI). Patients & methods: In total, 46 patients with HFrEF and 20 controls underwent metabolomic profiling, including liquid/gas chromatography–mass spectrometry and solid-phase microextraction volatilomics in plasma and urine. HFrEF was defined using left ventricular (LV) global longitudinal strain, EF and N-terminal pro hormone BNP. AECG and Echo AI were performed over 5 min, with a subset of patients undergoing a virtual reality mental stress test. Results: A-ECG had similar diagnostic accuracy as N-terminal pro hormone BNP for HFrEF (area under the curve = 0.95, 95% CI: 0.85–0.99), and correlated with global longitudinal strain (r = -0.77, p < 0.0001), while Echo AI-generated measurements correlated well with manually measured LV end diastolic volume r = 0.77, LV end systolic volume r = 0.8, LVEF r = 0.71, indexed left atrium volume r = 0.71 and indexed LV mass r = 0.6, p < 0.005. AI-LVEF and other HFrEF biomarkers had a similar discrimination for HFrEF (area under the curve AI-LVEF = 0.88; 95% CI: -0.03 to 0.15; p = 0.19). Virtual reality mental stress test elicited arrhythmic biomarkers on AECG and indicated blunted autonomic responsiveness (alpha 2 of RR interval variability, p = 1 × 10-4) in HFrEF. Conclusion: Multiomics-related machine learning shows promise for the assessment of HF.

scholarly journals Clinical outcomes of chemotherapy-induced subclinical left ventricular systolic dysfunction among breast cancer patients detected using echocardiographic global longitudinal strain

2020 ◽  
Vol 41 (Supplement_2) ◽  
Author(s):  
J.G.G Manaloto ◽  
M.K Cruz-Tan ◽  
R.H Tiongco ◽  
R.M Jimenez ◽  
G.H Cornelio
Keyword(s):  
Breast Cancer ◽  
Heart Failure ◽  
Cancer Patients ◽  
Clinical Outcomes ◽  
Longitudinal Strain ◽  
Systolic Dysfunction ◽  
Global Longitudinal Strain ◽  
Left Ventricular ◽  
Breast Cancer Patients ◽  
All Cause Mortality

Abstract Background Echocardiographic global longitudinal strain (GLS) detects early subclinical left ventricular (LV) systolic dysfunction, before the occurrence of a decreased LV ejection fraction. However, our local data is lacking to determine its impact to clinical outcomes. Purpose The study aimed to determine the clinical outcomes of breast cancer patients who developed subclinical LV systolic dysfunction as determined by an abnormal GLS post-chemotherapy. Methods This retrospective cohort study included 99 breast cancer patients who underwent anthracycline and/or HER-2 receptor inhibitor chemotherapy from January 1, 2016 to December 31, 2018 in a single tertiary hospital. Clinical outcomes of all-cause mortality and overt heart failure were compared between those with normal and abnormal GLS post-chemotherapy. Results The prevalence of subclinical LV systolic dysfunction was 18%, wherein 28% of them had subsequent overt heart failure, and 33% expired. Abnormal GLS occurred at a mean 3.5 months (range 1–8 months) after initiation of chemotherapy and at 8 months (range 6–10 months) after the entire chemotherapy sessions. Development into heart failure was observed at a mean of 6.7 months (range 4–12 months) after occurrence of abnormal GLS. Hypertension and age &gt;56 years were determined to be risk factors. Beta-blockers, ACE inhibitors and statins seemed to be non-protective in our cohort. Abnormalities in GLS were observed at a mean dose of 260 mg/m2 of epirubicin, lower than the dose described as high risk in the literature (600 mg/m2 for epirubicin). In trastuzumab, abnormal GLS occurred as early as 1 month after initiation. LVEF had no significant change within 2 months (p=0.56), but was significantly lower within 12 months post-chemotherapy (p=0.005). All-cause mortality was 3-fold higher (RR=3.00; p=0.02), and the risk to develop heart failure was 4 times higher (RR=4.74; p=0.008) in those with abnormal GLS. Conclusion The development of abnormal GLS post-chemotherapy was associated with subsequent development of overt heart failure and increased all-cause mortality. Abnormal GLS occurred at lower doses of epirubicin and as early as 1 month after initiating trastuzumab. We recommend echo surveillance with GLS monitoring beginning &gt;250 mg/m2 with anthracycline (and after 1–2 months of Trastuzumab), and to repeat at 1–2 months and 9–12 months post-chemotherapy. Funding Acknowledgement Type of funding source: None

scholarly journals Positive effect of dapagliflozin on left ventricular longitudinal function for type 2 diabetic mellitus patients with chronic heart failure

2020 ◽  
Vol 19 (1) ◽  
Author(s):  
Hidekazu Tanaka ◽  
Fumitaka Soga ◽  
Kazuhiro Tatsumi ◽  
Yasuhide Mochizuki ◽  
Hiroyuki Sano ◽  
...  
Keyword(s):  
Heart Failure ◽  
Ejection Fraction ◽  
Diastolic Function ◽  
Longitudinal Strain ◽  
Myocardial Function ◽  
Global Longitudinal Strain ◽  
Sglt2 Inhibitor ◽  
Left Ventricular ◽  
Lv Diastolic Function

Abstract Background The effect of sodium glucose cotransporter type 2 (SGLT2) inhibitor on left ventricular (LV) longitudinal myocardial function in type 2 diabetes mellitus (T2DM) patients with heart failure (HF) has remained unclear. Methods We analyzed data from our previous prospective multicenter study, in which we investigated the effect of the SGLT2 inhibitor dapagliflozin on LV diastolic functional parameters of T2DM patients with stable HF at five institutions in Japan. Echocardiography was performed at baseline and 6 months after administration of dapagliflozin. LV diastolic function was defined as the ratio of mitral inflow E to mitral e′ annular velocities (E/e′). LV longitudinal myocardial function was assessed as global longitudinal strain (GLS), which in turn was determined as the averaged peak longitudinal strain from standard LV apical views. Results E/e′ significantly decreased from 9.3 to 8.5 cm/s 6 months after administration of dapagliflozin (p = 0.020) as previously described, while GLS showed significant improvement from 15.5 ± 3.5% to 16.9 ± 4.1% (p < 0.01) 6 months after administration of dapagliflozin. Furthermore, improvement of GLS in HF with preserved ejection fraction patients was more significant from 17.0 ± 1.9% to 18.7 ± 2.0% (p < 0.001), compared to that in HF with mid-range ejection fraction and HF with reduced ejection fraction patients from 14.4 ± 2.4% to 15.5 ± 1.8% (p = 0.06) and from 8.1 ± 1.5% to 7.8 ± 2.1% (p = 0.44), respectively. It was noteworthy that multiple regression analysis showed that the change in GLS after administration of dapagliflozin was the only independent determinant parameters for the change in E/e′ after administration of dapagliflozin. Conclusion Dapagliflozin was found to be associated with improvement of LV longitudinal myocardial function, which led to further improvement of LV diastolic function of T2DM patients with stable HF. GLS-guided management may thus lead to improved management of T2DM patients with stable HF.

Sign in / Sign up

Export Citation Format

Share Document