Fully automated quantification of cardiac chamber and function assessment in 2-D echocardiography: clinical feasibility of deep learning-based algorithms

Deep Learning in Echocardiography for Enhanced Detection of Left Ventricular Function and Wall Motion Abnormalities

Cardiovascular diseases (CVDs) remain a leading cause of mortality worldwide, underscoring the need for advancements in diagnostic methodologies to improve early detection and treatment outcomes. This systematic review examines the integration of advanced deep learning (DL) techniques in echocardiography for detecting cardiovascular abnormalities, adhering to PRISMA 2020 guidelines. Through a comprehensive search across databases like IEEE Xplore, PubMed, and Web of Science, 29 studies were identified and analyzed, focusing on deep convolutional neural networks (DCNNs) and their role in enhancing the diagnostic precision of echocardiographic assessments. The findings highlight DL's capability to improve the accuracy and reproducibility of detecting and classifying echocardiographic data, particularly in measuring left ventricular function and identifying wall motion abnormalities. Despite these advancements, challenges such as data diversity, image quality, and the computational demands of DL models hinder their broader clinical adoption. In conclusion, DL offers significant potential to enhance the diagnostic capabilities of echocardiography. However, successful clinical implementation requires addressing issues related to data quality, computational demands, and system integration.

Integrating artificial intelligence into an echocardiography department: Feasibility and comparative study of automated versus human measurements in a high-volume clinical setting

BACKGROUND

Echocardiography is an important diagnostic tool in cardiology as it is essential for heart disease treatment. However, its time-consuming nature and reliance on user expertise constitutes a challenge for its use in high-volume clinics. Artificial intelligence (AI) offers the potential to automate tasks performed manually by echocardiographers and promises to improve efficiency and diagnostic consistency.

AIMS

To evaluate the integration of AI-based tools in a high-volume echocardiography department and assess the concordance of AI-generated measurements with manually-performed measurements.

METHODS

The study was conducted in the echocardiography department of Bordeaux University Hospital. Over 2months, 894 echocardiograms were performed by operators with three experience levels (nurses, residents and experts), with measurements performed by AI and humans. The statistical analyses assessed measurement agreement between both.

RESULTS

The AI system was successfully integrated into the hospital's infrastructure within 6weeks. Concordance analysis revealed good to very good agreement between AI and human measurements for most parameters, especially for ejection fraction (intraclass correlation coefficient [ICC]: 0.81, 95% confidence interval [95% CI]: 0.78-0.85) and Doppler-based flow measurements (mitral E wave velocity: ICC 0.97, 95% CI 0.95-0.98). Bland-Altman analysis showed a global mean difference of -4% with a standard deviation of 15%. Subgroup analysis revealed higher concordance for experts and residents compared with nurses (mean ICCs: 0.78 and 0.79 vs. 0.72, respectively).

CONCLUSION

AI can be effectively integrated into clinical echocardiography practice, with high agreement between AI and human measurements. Further research is needed to investigate the long-term impact on clinical outcomes and efficiency.

Evaluation of state-of-the-art deep learning models in the segmentation of the left and right ventricles in parasternal short-axis echocardiograms

Purpose

Previous studies on echocardiogram segmentation are focused on the left ventricle in parasternal long-axis views. Deep-learning models were evaluated on the segmentation of the ventricles in parasternal short-axis echocardiograms (PSAX-echo). Segmentation of the ventricles in complementary echocardiogram views will allow the computation of important metrics with the potential to aid in diagnosing cardio-pulmonary diseases and other cardiomyopathies. Evaluating state-of-the-art models with small datasets can reveal if they improve performance on limited data.

Approach

PSAX-echo was performed on 33 volunteer women. An experienced cardiologist identified end-diastole and end-systole frames from 387 scans, and expert observers manually traced the contours of the cardiac structures. Traced frames were pre-processed and used to create labels to train two domain-specific (Unet-Resnet101 and Unet-ResNet50), and four general-domain [three segment anything (SAM) variants, and the Detectron2] deep-learning models. The performance of the models was evaluated using the Dice similarity coefficient (DSC), Hausdorff distance (HD), and difference in cross-sectional area (DCSA).

Results

The Unet-Resnet101 model provided superior performance in the segmentation of the ventricles with 0.83, 4.93 pixels, and 106    pixel 2 on average for DSC, HD, and DCSA, respectively. A fine-tuned MedSAM model provided a performance of 0.82, 6.66 pixels, and 1252    pixel 2 , whereas the Detectron2 model provided 0.78, 2.12 pixels, and 116    pixel 2 for the same metrics, respectively.

Conclusions

Deep-learning models are suitable for the segmentation of the left and right ventricles in PSAX-echo. We demonstrated that domain-specific trained models such as Unet-ResNet provide higher accuracy for echo segmentation than general-domain segmentation models when working with small and locally acquired datasets.

Right Ventricular Pressure Waveform Analysis-Clinical Relevance and Future Directions

Continuous measurement of pressure in the right atrium and pulmonary artery has commonly been used to monitor right ventricular function in critically ill and surgical patients. This approach is largely based upon the assumption that right atrial and pulmonary arterial pressures provide accurate surrogates for diastolic filling and peak right ventricular pressures, respectively. However, due to both technical and physiologic factors, this assumption is not always true. Accordingly, recent studies have begun to emphasize the potential clinical value of also measuring right ventricular pressure at the bedside. This has highlighted both past and emerging research demonstrating the utility of analyzing not only the amplitude of right ventricular pressure but also the shape of the pressure waveform. This brief review summarizes data demonstrating that combining conventional measurements of right ventricular pressure with variables derived from waveform shape allows for more comprehensive and ideally continuous bedside assessment of right ventricular function, particularly when combined with stroke volume measurement or 3D echocardiography, and discusses the potential use of right ventricular pressure analysis in computational models for evaluating cardiac function.

Artificial intelligence-enhanced automation for M-mode echocardiographic analysis: ensuring fully automated, reliable, and reproducible measurements

To enhance M-mode echocardiography's utility for measuring cardiac structures, we developed and evaluated an artificial intelligence (AI)-based automated analysis system for M-mode images through the aorta and left atrium [M-mode (Ao-LA)], and through the left ventricle [M-mode (LV)]. Our system, integrating two deep neural networks (DNN) for view classification and image segmentation, alongside an auto-measurement algorithm, was developed using 5,958 M-mode images [3,258 M-mode (LA-Ao), and 2,700 M-mode (LV)] drawn from a nationwide echocardiographic dataset collated from five tertiary hospitals. The performance of view classification and segmentation DNNs were evaluated on 594 M-mode images, while automatic measurement accuracy was tested on separate internal test set with 100 M-mode images as well as external test set with 280 images (140 sinus rhythm and 140 atrial fibrillation). Performance evaluation showed the view classification DNN's overall accuracy of 99.8% and segmentation DNN's Dice similarity coefficient of 94.3%. Within the internal test set, all automated measurements, including LA, Ao, and LV wall and cavity, resonated strongly with expert evaluations, exhibiting Pearson's correlation coefficients (PCCs) of 0.81-0.99. This performance persisted in the external test set for both sinus rhythm (PCC, 0.84-0.98) and atrial fibrillation (PCC, 0.70-0.97). Notably, automatic measurements, consistently offering multi-cardiac cycle readings, showcased a stronger correlation with the averaged multi-cycle manual measurements than with those of a single representative cycle. Our AI-based system for automatic M-mode echocardiographic analysis demonstrated excellent accuracy, reproducibility, and speed. This automated approach has the potential to improve efficiency and reduce variability in clinical practice.

Deep learning supported echocardiogram analysis: A comprehensive review

An echocardiogram is a sophisticated ultrasound imaging technique employed to diagnose heart conditions. The transthoracic echocardiogram, one of the most prevalent types, is instrumental in evaluating significant cardiac diseases. However, interpreting its results heavily relies on the clinician's expertise. In this context, artificial intelligence has emerged as a vital tool for helping clinicians. This study critically analyzes key state-of-the-art research that uses deep learning techniques to automate transthoracic echocardiogram analysis and support clinical judgments. We have systematically organized and categorized articles that proffer solutions for view classification, enhancement of image quality and dataset, segmentation and identification of cardiac structures, detection of cardiac function abnormalities, and quantification of cardiac functions. We compared the performance of various deep learning approaches within each category, identifying the most promising methods. Additionally, we highlight limitations in current research and explore promising avenues for future exploration. These include addressing generalizability issues, incorporating novel AI approaches, and tackling the analysis of rare cardiac diseases.

Enhancing Arrhythmogenic Right Ventricular Cardiomyopathy Detection and Risk Stratification: Insights from Advanced Echocardiographic Techniques

INTRODUCTION

The echocardiographic diagnosis criteria for arrhythmogenic right ventricular cardiomyopathy (ARVC) are highly specific but sensitivity is low, especially in the early stages of the disease. The role of echocardiographic strain in ARVC has not been fully elucidated, although prior studies suggest that it can improve the detection of subtle functional abnormalities. The purposes of the study were to determine whether these advanced measures of right ventricular (RV) dysfunction on echocardiogram, including RV strain, increase diagnostic value for ARVC disease detection and to evaluate the association of echocardiographic parameters with arrhythmic outcomes.

METHODS

The study included 28 patients from the Heart Institute of São Paulo ARVC cohort with a definite diagnosis of ARVC established according to the 2010 Task Force Criteria. All patients were submitted to ECHO's advanced techniques including RV strain, and the parameters were compared to prior conventional visual ECHO and CMR.

RESULTS

In total, 28 patients were enrolled in order to perform ECHO's advanced techniques. A total of 2/28 (7%) patients died due to a cardiovascular cause, 2/28 (7%) underwent heart transplantation, and 14/28 (50%) patients developed sustained ventricular arrhythmic events. Among ECHO's parameters, RV dilatation, measured by RVDd (p = 0.018) and RVOT PSAX (p = 0.044), was significantly associated with arrhythmic outcomes. RV free wall longitudinal strain < 14.35% in absolute value was associated with arrhythmic outcomes (p = 0.033).

CONCLUSION

Our data suggest that ECHO's advanced techniques improve ARVC detection and that abnormal RV strain can be associated with arrhythmic risk stratification. Further studies are necessary to better demonstrate these findings and contribute to risk stratification in ARVC, in addition to other well-known risk markers.