Fully Automated Echocardiogram Interpretation in Clinical Practice

Zhang, Jeffrey; Gajjala, Sravani; Agrawal, Pulkit; Tison, Geoffrey H.; Hallock, Laura A.; Beussink‐Nelson, Lauren; Lassen, Mats Christian Højbjerg; Fan, Eugene; Aras, Mandar A.; Jordan, Cha Randle; Fleischmann, Kirsten E.; Melisko, Michelle; Qasim, Atif; Shah, Sanjiv J.; Bajcsy, Růžena; Deo, Rahul C.

doi:10.1161/circulationaha.118.034338

Cited by 671 publications

(600 citation statements)

References 25 publications

Supporting

Mentioning

572

Contrasting

Unclassified

Order By: Relevance

“…Knackstedt et al demonstrated that the left ventricular ejection fraction and longitudinal strain could be accurately and reproducibly computed from echo data using ML‐enabled software 17 . Zhang et al trained convolutional neural networks to obtain measurements of the left ventricle and predict various disease states including hypertrophic cardiomyopathy, cardiac amyloidosis, and pulmonary arterial hypertension 8 . While these studies demonstrate feasibility for utilization of ML approaches for LV functional assessment, there are limited data for its use for RV assessment, likely owing to the geometric complexity of RV structure.…”

Section: Discussionmentioning

confidence: 99%

“…Machine learning (ML)–based methodologies have the potential to provide fully automated image analysis. Conventional neural networks–based segmentation techniques have been applied to echo, though with focus primarily on left ventricular chamber size and systolic function quantification 8 . While a recent study examined a ML approach for three‐dimensional echocardiography (3DE) assessment of RV volume and EF, 9 limited clinical availability of 3DE is a known barrier for widespread utilization.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Development of novel machine learning model for right ventricular quantification on echocardiography—A multimodality validation study

et al. 2020

View full text Add to dashboard Cite

Purpose: Echocardiography (echo) is widely used for right ventricular (RV) assessment. Current techniques for RV evaluation require additional imaging and manual analysis; machine learning (ML) approaches have the potential to provide efficient, fully automated quantification of RV function.Methods: An automated ML model was developed to track the tricuspid annulus on echo using a convolutional neural network approach. The model was trained using 7791 image frames, and automated linear and circumferential indices quantifying annular displacement were generated. Automated indices were compared to an independent reference of cardiac magnetic resonance (CMR) defined RV dysfunction (RVEF < 50%). Results:A total of 101 patients prospectively underwent echo and CMR: Fully automated annular tracking was uniformly successful; analyses entailed minimal processing time (<1 second for all) and no user editing. Findings demonstrate all automated annular shortening indices to be lower among patients with CMR-quantified RV dysfunction (all P < .001). Magnitude of ML annular displacement decreased stepwise in relation to population-based tertiles of TAPSE, with similar results when ML analyses were localized to the septal or lateral annulus (all P ≤ .001). Automated segmentation techniques provided good diagnostic performance (AUC 0.69-0.73) in relation to CMR reference and compared to conventional RV indices (TAPSE and S′) with high negative predictive value (NPV 84%-87% vs 83%-88%). Reproducibility was higher for ML algorithm as compared to manual segmentation with zero inter-and intraobserver variability and ICC 1.0 (manual ICC: 0.87-0.91). Conclusions:This study provides an initial validation of a deep learning system for RV assessment using automated tracking of the tricuspid annulus. | 689BEECY Et al.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Development of novel machine learning model for right ventricular quantification on echocardiography—A multimodality validation study

et al. 2020

View full text Add to dashboard Cite

show abstract

“…?B, a selection of the most common standard echocardiogram views were evaluated for model performance. Images from each study were classified using a previously described supervised training method 5 . We sought to identify the most information-rich views by training separate models on the subsets of dataset images of only one cardiac view.…”

Section: /14mentioning

confidence: 99%

“…Machine learning on echocardiography images can streamline repetitive tasks in the clinical workflow, standardize interpretation in areas with insufficient qualified cardiologists, and more consistently produce echocardiographic measurements.Related works Current literature have already shown that it is possible to identify standard echocardiogram views from unlabeled datasets. 5,6,24 Previous works have used convolutional neural networks (CNNs) trained on images and videos from echocardiography to perform segmentation to identify cardiac structures and derive cardiac function. In this study, we extend previous analyses to show that EchoNet, our deep learning model using echocardiography images, local cardiac structures and anatomy, estimate volumetric measurements and metrics of cardiac function, and predict systemic human phenotypes that modify cardiovascular risk.…”

mentioning

confidence: 99%

Deep Learning Interpretation of Echocardiograms

Ghorbani

Ouyang

Abid

et al. 2019

Preprint

View full text Add to dashboard Cite

Echocardiography uses ultrasound technology to capture high temporal and spatial resolution images of the heart and surrounding structures and is the most common imaging modality in cardiovascular medicine. Using convolutional neural networks on a large new dataset, we show that deep learning applied to echocardiography can identify local cardiac structures, estimate cardiac function, and predict systemic phenotypes that modify cardiovascular risk but not readily identifiable to human interpretation. Our deep learning model, EchoNet, accurately identified the presence of pacemaker leads (AUC = 0.89), enlarged left atrium (AUC = 0.85), normal left ventricular wall thickness (AUC = 0.75), left ventricular end systolic and diastolic volumes(R 2 = 0.73 and R 2 = 0.68), and ejection fraction (R 2 = 0.48) as well as predicted systemic phenotypes of age (R 2 = 0.46), sex (AUC = 0.88), weight (R 2 = 0.56), and height (R 2 = 0.33). Interpretation analysis validates that EchoNet shows appropriate attention to key cardiac structures when performing human-explainable tasks and highlight hypothesis-generating regions of interest when predicting systemic phenotypes difficult for human interpretation. Machine learning on echocardiography images can streamline repetitive tasks in the clinical workflow, standardize interpretation in areas with insufficient qualified cardiologists, and more consistently produce echocardiographic measurements.Related works Current literature have already shown that it is possible to identify standard echocardiogram views from unlabeled datasets. 5,6,24 Previous works have used convolutional neural networks (CNNs) trained on images and videos from echocardiography to perform segmentation to identify cardiac structures and derive cardiac function. In this study, we extend previous analyses to show that EchoNet, our deep learning model using echocardiography images, local cardiac structures and anatomy, estimate volumetric measurements and metrics of cardiac function, and predict systemic human phenotypes that modify cardiovascular risk. Additionally, we show the first application of interpretation frameworks to understand deep learning models from echocardiogram images. Human identifiable features, such as the presence of pacemaker and defibrillator leads, left ventricular hypertrophy, and abnormal left atrial chamber size identified by our convolutional neural network were validated using interpretation frameworks to highlight the most relevant regions of interest. To the best of our knowledge, we develop the first deep learning model that can directly predict age, sex, weight and height from echocardiogram images and use interpretation methods to understand how the model predicts these systemic phenotypes difficult for human interpreters. 2/14Lessons from model training and experiments EchoNet performance greatly improved with efforts to augment data size, homogenize input data, and with optimize model training with hyperparameter search. Our experience shows that increasing number of unique patien...

show abstract

“…This would reduce inter-and intraobserver variability and create a unique quantification system in order to standardize diagnostic and monitoring scores. Such methodology, supported by artificialintelligence software, has been successfully tested for other ultrasound automated techniques [7]. The potential advantages in terms of faster data collection without increased costs and patients risks are intuitive.…”

mentioning

confidence: 99%

Lung ultrasound and B-lines quantification inaccuracy: B sure to have the right solution

et al. 2020

View full text Add to dashboard Cite

Dear Editor,We read with great interest the letter form Haaksma and co-workers titled "Lung ultrasound and B-lines: B careful!" [1]. The authors found that lung ultrasound (LUS) reproducibility of B-lines detection with different transducers and raters was poor to moderate and raised a relevant issue.B-lines are dynamic LUS artifacts, moving and potentially changing appearance over the respiratory cycle, associated with increased extravascular lung water or partial lung loss of aeration. Although the recognition of B-lines and discrimination of A-from B-pattern are simple tasks, the quantification of B-lines and the assessment of their spacing can be challenging: easily counted when few, it becomes impossible to enumerate them precisely when numerous and tending to merge (as often happens in the interstitial-alveolar syndromes).Semiquantitative methods have been proposed to quantify B-lines based on visual estimation of screen percentage occupied by them [2] or on the presence/absence of their coalescence [3]. These methods may be prone to errors either due to inter-operator "eyeballing" variability or to assessment of their coalescence without considering its overall pleural extension.The current recommendation to assess the percentage of pleural line occupied by B-lines (rather than counting their maximum number over each ultrasound scan) [2] may also lead to inaccurate results.At present, no tool is available to quantify the percentage of pleural line occupied by B-lines. This may not be a simple cognitive process because: (1) the distance

show abstract

Fully Automated Echocardiogram Interpretation in Clinical Practice

Abstract: Supplemental Digital Content is available in the text.

Cited by 671 publications

References 25 publications

Development of novel machine learning model for right ventricular quantification on echocardiography—A multimodality validation study

Development of novel machine learning model for right ventricular quantification on echocardiography—A multimodality validation study

Deep Learning Interpretation of Echocardiograms

Lung ultrasound and B-lines quantification inaccuracy: B sure to have the right solution

Contact Info

Product

Resources

About