Transmitral inflow wave and progression from paroxysmal to permanent atrial fibrillation in Asian people

Machine Learning Modeling to Predict Atrial Fibrillation Detection in Embolic Stroke of Undetermined Source Patients

BACKGROUND

In patients with embolic stroke of undetermined source (ESUS), occult atrial fibrillation (AF) has been implicated as a key source of cardioembolism. However, only a minority acquire implantable cardiac loop recorders (ILRs) to detect occult paroxysmal AF, partly due to financial cost and procedural inconvenience. Without the initiation of appropriate anticoagulation, these patients are at risk of increased ischemic stroke recurrence. Hence, cost-effective and accurate methods of predicting AF in ESUS patients are highly sought after.

OBJECTIVE

We aimed to incorporate clinical and echocardiography data into machine learning (ML) algorithms for AF prediction on ILRs in ESUS.

METHODS

This was a single-center cohort study that included 157 consecutive patients diagnosed with ESUS from October 2014 to October 2017 who had ILR evaluation. We developed four ML models, with hyperparameters tuned, to predict AF detection on an ILR.

RESULTS

The median age of the cohort was 67 (IQR 59-74) years old and the median monitoring duration was 1051 (IQR 478-1287) days. Of the 157 patients, 32 (20.4%) had occult AF detected on the ILR. Support vector machine predicted for AF with a 95% confidence interval area under the receiver operating characteristic curve (AUC) of 0.736-0.737, multilayer perceptron with an AUC of 0.697-0.708, XGBoost with an AUC of 0.697-0.697, and random forest with an AUC of 0.663-0.674. ML feature importance found that age, HDL-C, and admitting heart rate were important non-echocardiography variables, while peak mitral A-wave velocity and left atrial volume were important echocardiography parameters aiding this prediction.

CONCLUSION

Machine learning modeling incorporating clinical and echocardiographic variables predicted AF in ESUS patients with moderate accuracy.

A Review of the Role of Transthoracic and Transesophageal Echocardiography, Computed Tomography, and Magnetic Resonance Imaging in Cardioembolic Stroke

Stroke is a major source of morbidity and mortality worldwide, accounting for the second largest cause of mortality and the third greatest cause of disability. Stroke is frequently preceded by a transient ischemic attack (TIA). The etiologies of 20-30% of ischemic strokes are unknown, and thus are termed "cryptogenic strokes". About 25% of ischemic strokes are cardioembolic. Strokes occur at a rate of around 2% per year in individuals with heart failure with reduced ejection fraction (HFrEF), with a strong correlation between stroke risk and the degree of ventricular impairment. Furthermore, stroke risk is augmented in the absence of anticoagulation therapy. Cardioembolic strokes, when treated inadequately, have a greater predilection for recurrences than atherothrombotic strokes, both early and late in life. The role of a patent foramen ovale in strokes, specifically in "cryptogenic strokes", is a matter of concern that deserves due attention. The use of tissue-engineered heart valves and aspirin for minimizing the risk of stroke is recommended. Transthoracic echocardiography (TTE) is advantageous for assessing heart function in the acute phase of ischemic stroke. Transesophageal echocardiography (TEE) is considered the criterion standard procedure for detecting LAA thrombi. Computed tomography (CT) scans are good imaging modalities for identifying and excluding bleeding. Magnetic resonance imaging (MRI) images are by far the most effective imaging technique available for assessing the brain parenchymal state. We conducted a thorough review of the literature on the use of imaging modalities, highlighting the important contribution of TTE, TEE, CT, and MRI in the evaluation of cardioembolic stroke.

Left atrial appendage shape impacts on the left atrial flow hemodynamics: A numerical hypothesis generating study on two cases

BACKGROUND AND OBJECTIVES

The left atrial appendage (LAA) is the most common region for thrombus formation in atrial fibrillation (AF). Morphological parameters such as shape, size, and LAA volume can cause insufficient effectiveness of available therapeutic options. This study aimed to examine blood flow inside LAA and its removal effects. Computational fluid dynamic (CFD) simulations were carried out on two patients with different morphologies.

METHODS

Two patients' CT was used to reconstruct the 3D geometries of the left atrium (LA) and left atrial appendage (LAA). Then, the geometries were refined in the mentioned software, and the LAA in some models was removed. Next, in generated 3D volume mesh, sinus rhythm (SR) and atrial fibrillation (AF) outflow velocity were imposed at the mitral valve as boundary conditions. Finally, CFD simulation was conducted to analyzing blood flow within LA with/without LA.

RESULTS

The results confirmed that velocity and vorticity decreased under AF conditions inside the LA domain for both patients. However, removing LAA may cause unpredictable consequences, due to different shape and volume of LAA. LAA removal had insignificant effects on velocity and vorticity within LA in SR-mitral outflow. However, removing LAA increased the blood flow rate by 9.15% and vorticity by 7.27% for patient one under AF rhythm (SR)-outflow. In contrast, for patient two, LAA removal in both AF and SR decreased velocity and vorticity within the LA domain. In SR-mitral outflow, velocity dropped by 18.8 %, and vorticity by 13.2%. Also, under AF velocity and vorticity decreased by 23.33% and 18.6% respectively. Meanwhile, the results indicated that the vorticity magnitude increased inside the LAA under AF associated with the risk of thrombus formation, particularly for patient one under AF. The distal part of LAA in both patients was the most common region for blood stasis because of the lowest velocity magnitude.

CONCLUSION

Overall, the morphology of LAA could be the critical parameter to determine the possibility of thrombosis formation, particularly under AF conditions. High volume, low blood flow velocity and two-lobe-appendage are more likely to have blood stasis. Furthermore, the morphology difference can affect the LAA removal result and make it more complicated. So, it could be challenging to generalize LAA removal as a therapeutic option for different patients. The implication of this CFD observation needs more investigation.

Echocardiographic predictors of cardioembolic stroke due to underlying atrial fibrillation: Reliable left atrial remodeling signs in acute stroke

INTRODUCTION

Atrial remodeling due to high-burden atrial fibrillation (AF) is associated with cardioembolic stroke (CES). As not all CESs is caused by AF, we analyzed the diagnostic values of each echocardiographic parameter to distinguish likely AF-related CES in acute stroke patients while in non-AF rhythm.

METHODS

The data of consecutive patients with acute ischemic stroke in sinus rhythm between 2012 and 2015 were obtained. The echocardiographic parameters of patients with CES due to underlying AF (n = 61) and control patients (n = 319) with either large artery atherosclerosis or small-vessel occlusion were compared using receiver operating characteristic curves and logistic regression analyses. Each parameter was reassessed in acute stroke patients through a validation study using the same database with different periods of generalization.

RESULTS

CES patients with underlying AF showed a significantly larger left atrial volume index (LAVi), higher mitral inflow E wave (E), and lower A wave (A) than the controls. The area under the curve (AUC) (95% confidence interval) for diagnosing CES due to underlying AF was significantly higher for LAVi/A than for LAVi (0.785 versus 0.696, P < 0.01). Among patients aged >60 years, the E/A ratio had the highest AUC (0.857) of the parameters. The cut-off values were ≥ 0.70 (sensitivity, 55.7%; specificity, 90.9%) and ≥ 0.82 (sensitivity, 71.4%; specificity, 84.1%) for LAVi/A and the E/A ratio, respectively, in patients >60 years. The cut-off values of all parameters showed similar trends in a validation study.

CONCLUSION

LAVi/A is a useful indicator for distinguishing CES patients with underlying AF regardless of age, and the E/A ratio is reliable among patients aged >60 years in evaluation during acute stroke admission.