scholarly journals Fully automated global longitudinal strain assessment using artificial intelligence developed and validated by a UK-wide echocardiography expert collaborative

2021 ◽  
Vol 42 (Supplement_1) ◽  
Author(s):  
C Stowell ◽  
J Howard ◽  
C Demetrescu ◽  
S Bhattacharyya ◽  
K Mangion ◽  
...  
Keyword(s):  
Open Source ◽  
Longitudinal Strain ◽  
Left Ventricular ◽  
Expert Consensus ◽  
Validation Dataset ◽  
Visual Grading ◽  
Online Platform ◽  
Longitudinal Function ◽  
Median Error ◽  
The Individual

Abstract Background Left ventricular longitudinal strain has been reported to deliver reproducibility, sensitivity and prognostic value over and above ejection fraction. However, it currently relies on uninspectable proprietary algorithms and suffers from a lack of widespread clinical use. Uptake may be improved by increasing user trust through greater transparency. Purpose We therefore developed a machine-learning based method, trained, and validated with accredited experts from our AI Echocardiography Collaborative. We make the dataset, code, and trained network freely available under an open-source license. Methods AI enables strain to be calculated without relying on speckle tracking by directly locating key points and borders across frames. Strain can then be calculated as the fractional shortening of the left ventricular perimeter. We first curated a dataset of 7523 images, including 2587 apical four chamber, each labelled by a single expert from our collaboration of 17 hospitals, using our online platform (Figure 1). Using both this dataset and a semi-supervised approach, we trained a 3d convolutional neural network to identify the annulus, apex, and the endocardial border throughout the cardiac cycle. Separately, we constructed an external validation dataset of 100 apical 4 chamber video-loops. The systolic and diastolic frame were identified, and each image was separately labelled by 11 experts. From these labels we then derived the expert consensus strain for each of the 100 video loops. These experts also ordered all 100 echocardiograms by their visual grading of left ventricular longitudinal function. Finally, a single expert calculated strain using two different proprietary commercial packages (A and B). Results Consensus strain measurements (obtained by averaging individual assessments by the 11 experts) across the 100 cases ranged from −4% to −27%, with strong correlations with the individual experts and machine methods (Figure 2). Using each cases' consensus across experts as the gold standard, median error from consensus was 3.1% for individual experts, 3.4% for Propriety A, 2.6% for Proprietary B, 2.6% for our AI. Using the visual grading of longitudinal strain as the reference, the 11 individual experts and 4 machine methods each showed significant correlation: coefficients ranged from 0.55 to 0.69 for experts, and for Proprietary A was 0.68, Proprietary B 0.69, and our AI 0.69. Conclusions Our open-source, vendor-independent AI-based strain measure automatically produces values that agree with expert consensus, as strongly as the individual experts do. It also agrees with the subjective visual ranking by longitudinal function. Our open-source AI strain performs at least as well as closed-source speckle-based approaches, and may enable increased clinical and research use of longitudinal strain. FUNDunding Acknowledgement Type of funding sources: Public grant(s) – National budget only. Main funding source(s): NIHR Imperial BRC ITMAT.Dr Howard was additionally funded by Wellcome. Figure 1. Collaborative online platform Figure 2. Correlations between strain methods

scholarly journals Automated Left Ventricular Dimension Assessment Using Artificial Intelligence Developed and Validated by a UK-Wide Collaborative

2021 ◽  
Author(s):  
James P. Howard ◽  
Catherine C. Stowell ◽  
Graham D. Cole ◽  
Kajaluxy Ananthan ◽  
Camelia D. Demetrescu ◽  
...  
Keyword(s):  
Artificial Intelligence ◽  
Correlation Coefficient ◽  
Intraclass Correlation Coefficient ◽  
Interventricular Septum ◽  
Intraclass Correlation ◽  
Left Ventricular ◽  
Expert Consensus ◽  
Training Dataset ◽  
Validation Dataset ◽  
Online Platform

Background: Artificial intelligence (AI) for echocardiography requires training and validation to standards expected of humans. We developed an online platform and established the Unity Collaborative to build a dataset of expertise from 17 hospitals for training, validation, and standardization of such techniques. Methods: The training dataset consisted of 2056 individual frames drawn at random from 1265 parasternal long-axis video-loops of patients undergoing clinical echocardiography in 2015 to 2016. Nine experts labeled these images using our online platform. From this, we trained a convolutional neural network to identify keypoints. Subsequently, 13 experts labeled a validation dataset of the end-systolic and end-diastolic frame from 100 new video-loops, twice each. The 26-opinion consensus was used as the reference standard. The primary outcome was precision SD, the SD of the differences between AI measurement and expert consensus. Results: In the validation dataset, the AI’s precision SD for left ventricular internal dimension was 3.5 mm. For context, precision SD of individual expert measurements against the expert consensus was 4.4 mm. Intraclass correlation coefficient between AI and expert consensus was 0.926 (95% CI, 0.904–0.944), compared with 0.817 (0.778–0.954) between individual experts and expert consensus. For interventricular septum thickness, precision SD was 1.8 mm for AI (intraclass correlation coefficient, 0.809; 0.729–0.967), versus 2.0 mm for individuals (intraclass correlation coefficient, 0.641; 0.568–0.716). For posterior wall thickness, precision SD was 1.4 mm for AI (intraclass correlation coefficient, 0.535 [95% CI, 0.379–0.661]), versus 2.2 mm for individuals (0.366 [0.288–0.462]). We present all images and annotations. This highlights challenging cases, including poor image quality and tapered ventricles. Conclusions: Experts at multiple institutions successfully cooperated to build a collaborative AI. This performed as well as individual experts. Future echocardiographic AI research should use a consensus of experts as a reference. Our collaborative welcomes new partners who share our commitment to publish all methods, code, annotations, and results openly.

scholarly journals Right ventricular dysfunction parallels left ventricular functional involvement in women with breast cancer experiencing subclinical cardiotoxicity

2021 ◽  
Vol 22 (Supplement_1) ◽  
Author(s):  
C Santoro ◽  
R Soloperto ◽  
O Casciano ◽  
R Esposito ◽  
M Lembo ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Heart Failure ◽  
Cancer Therapy ◽  
Right Ventricular ◽  
Longitudinal Strain ◽  
Global Longitudinal Strain ◽  
Left Ventricular ◽  
Breast Cancer Patients ◽  
Longitudinal Function ◽  
Oncologic Patients

Abstract Funding Acknowledgements Type of funding sources: None. Background Cancer therapy related cardiac toxicity disease (CRCTD) of the left ventricle (LV)can influence the outcome of oncologic patients. Little is known on CRCTD related right ventricular (RV)dysfunction even though RV involvement has been proven to be a remarkable prognosticator in heart failure. Purpose To analyse parallel changes in LV and RV function occurring during the course of cancer therapy in women affected by breast cancer by using both standard and speckle tracking echocardiography. Methods Fifty Her-2 positive breast cancer women (age = 53.6 ± 11.7 years) underwent sequential cancer therapy protocol including anthracycline (ANT) epirubicine + cyclophosphamide (4 cycles) followed by a total amount of 18 cycles with trastuzumab (TRZ) + paclitaxel. A complete echo-Doppler exam, including LV and RV global longitudinal strain (GLS)as well as RV septal and free wall longitudinal strain (SLS and FWLS respectively) assessment, was performed at baseline, after ANT end and after TRZ completion. Patients with overt heart failure and LV ejection fraction < 50%, coronary artery disease,atrial fibrillation, hemodinamically significant valve disease and inadequate echo were excluded. Overt CRCTD was defined according guidelines and both subclinical LV and RV CRCTD as a LV and RV GLS drop from baseline >15%. Results None of the patients experienced overt CTCRD but 6 patients (14%) showed subclinical LV dysfunction and 33 (66%) had a significant drop of RV longitudinal function.The comparison of standard echo-Doppler exam at baseline and after ANT and TRZ completion did not show significant changes of LV and RV systolic and diastolic parameters. Conversely, a progressive significant reduction of RV GLS (p < 0.002 after TRZ), SLS and FWLS and, with a lower extent, of LV GLS (p < 0.02 after TRZ) was observed after ANT and TRZ completion (Figure). Percentage reduction in RV GLS (DRV GLS) from baseline to ANT end correlated with LV GLS both at EC end (r=-0.40, p = 0.006) and after TRZ completion (r=-0.62, p < 0.0001). Conclusions Detrimental cardiac effects of cancer therapy involve both LV and RV systolic longitudinal function. Progressive RV dysfunction is evident through ANT and TRZ treatment. Early RV dysfunction parallels LV involvement and predicts subsequent LV subclinical dysfunction. A comprehensive LV and RV longitudinal function assessment might better predict the onset of CRCTD in breast cancer patients. Abstract Figure.

3077Load dependency of left ventricular longitudinal strain by 2D-speckle tracking echocardiography: myth or reality?

2019 ◽  
Vol 40 (Supplement_1) ◽  
Author(s):  
M J Nobre De Matos Pereira Vieira ◽  
D Ribeiro ◽  
N Craveiro ◽  
R Teixeira ◽  
L Pitta ◽  
...  
Keyword(s):  
Speckle Tracking ◽  
Longitudinal Strain ◽  
Speckle Tracking Echocardiography ◽  
Vena Cava ◽  
Left Atrial Volume ◽  
Left Ventricular ◽  
Healthy Individuals ◽  
Longitudinal Function ◽  
Maximal Diameter ◽  
2D Speckle Tracking

Abstract Background 2D-speckle tracking echocardiography (2D-STE) derived strain measurements has been proposed as a non-invasive measure of myocardial deformation and function. However, the effects of left ventricular (LV) loading conditions on 2D-STE derived LV longitudinal strain (GLS) have not been totally elucidated and the results of some studies regarding the load dependency of GLS are controversial. Purpose To characterize the effects of acute load change (preload increase) on LV GLS. Methods and results We evaluated the variation of LV GLS by 2D-STE, in response to a preload increasing maneuver (leg lifting maneuver – LLM), in a population of 30 healthy individuals. Clinical, demographic and echocardiographic parameters (including LV longitudinal mechanics obtained with 2D-STE before and after LLM) were described. The population had a mean age of 27±4 years and 73% were women. Increased preload to the heart with LLM was confirmed by an increase in the maximal diameter of the inferior vena cava (16±3.5 vs 22±3.3 mm, p<0.01). No significant changes in left atrial volume, LV ejection volume and LV ejection fraction were observed in response to the LLM. There was a significant variation of global LV GLS (−21.9±2.3 vs −23.2±1.6%, p<0.001, Δ 1.25%, 95% CI 0.5–1.91) – figure. An increase in right ventricular longitudinal function with LLM (TAPSE 22.5±5.4 vs 25.5±0.5 mm, p=0.005, Δ 2.9, 95% CI 0.9–4.8) was also observed. Conclusion To our knowledge this is the first study performed to assess the effect of preload increase in GLS using the LLM in healthy individuals. In this study, the absolute LV GLS value increased significantly in response to preload increase (LLM). The dependence of GLS on preload is in accordance with the Frank-Starling Law, in which an increase in preload in a healthy individual lead to an increase in myocardial contractility. These findings suggest that LV GLS is a sensitive parameter for detecting subtle changes in LV longitudinal function.

Right ventricular dysfunction parallels left ventricular functional involvement in women with breast cancer experiencing subclinical cardiotoxicity

2020 ◽  
Vol 41 (Supplement_2) ◽  
Author(s):  
C Santoro ◽  
R Soloperto ◽  
O Casciano ◽  
R Esposito ◽  
F Luciano ◽  
...  
Keyword(s):  
Breast Cancer ◽  
Heart Failure ◽  
Cancer Therapy ◽  
Right Ventricular ◽  
Longitudinal Strain ◽  
Global Longitudinal Strain ◽  
Left Ventricular ◽  
Breast Cancer Patients ◽  
Longitudinal Function ◽  
Oncologic Patients

Abstract Background Cancer therapy related cardiac toxicity disease (CRCTD) of the left ventricle (LV)can influence the outcome of oncologic patients. Little is known on CRCTD related right ventricular (RV)dysfunction even though RV involvement has been proven to be a remarkable prognosticator in heart failure. Purpose To analyse parallel changes in LV and RV function occurring during the course of cancer therapy in women affected by breast cancer by using both standard and speckle tracking echocardiography. Methods Fifty Her-2 positive breast cancer women (age = 53.6±11.7 years) underwent sequential cancer therapy protocol including anthracycline (ANT) epirubicine + cyclophosphamide (4 cycles) followed by a total amount of 18 cycles with trastuzumab (TRZ) + paclitaxel. A complete echo-Doppler exam, including LV and RV global longitudinal strain (GLS)as well as RV septal and free wall longitudinal strain (SLS and FWLS respectively) assessment, was performed at baseline, after ANT end and after TRZ completion. Patients with overt heart failure and LV ejection fraction &lt;50%, coronary artery disease,atrial fibrillation, hemodinamically significant valve disease and inadequate echo were excluded. Overt CRCTD was defined according guidelines and both subclinical LV and RV CRCTD as a LV and RV GLS drop from baseline &gt;15%. Results None of the patients experienced overt CTCRD but 6 patients (14%) showed subclinical LV dysfunction and 33 (66%) had a significant drop of RV longitudinal function.The comparison of standard echo-Doppler exam at baseline and after ANT and TRZ completion did not show significant changes of LV and RV systolic and diastolic parameters. Conversely, a progressive significant reduction of RV GLS (p&lt;0.002 after TRZ), SLS and FWLS and, with a lower extent, of LV GLS (p&lt;0.02 after TRZ) was observed after ANT and TRZ completion (Figure). Percentage reduction in RV GLS (DRV GLS) from baseline to ANT end correlated with LV GLS both at EC end (r=−0.40, p=0.006) and after TRZ completion (r=−0.62, p&lt;0.0001). Conclusions Detrimental cardiac effects of cancer therapy involve both LV and RV systolic longitudinal function. Progressive RV dysfunction is evident through ANT and TRZ treatment. Early RV dysfunction parallels LV involvement and predicts subsequent LV subclinical dysfunction. A comprehensive LV and RV longitudinal function assessment might better predict the onset of CRCTD in breast cancer patients. LV and RV strain during cancer therapy Funding Acknowledgement Type of funding source: None

scholarly journals Automated left ventricular dimension assessment using artificial intelligence

2021 ◽  
Vol 42 (Supplement_1) ◽  
Author(s):  
C Stowell ◽  
J Howard ◽  
G Cole ◽  
K Ananthan ◽  
C Demetrescu ◽  
...  
Keyword(s):  
Artificial Intelligence ◽  
External Validation ◽  
Posterior Wall ◽  
Left Ventricular ◽  
Validation Dataset ◽  
Left Ventricular Dimension ◽  
Collaborative Platform ◽  
Consensus Standard ◽  
The Individual ◽  
Ventricular Dimension

Abstract Background and purpose Artificial intelligence (AI) has the potential to greatly improve efficiency and reproducibility of quantification in echocardiography, but to gain widespread use it must both meet expert standards of excellence and have a transparent methodology. We developed an online platform to enable multiple collaborators to annotate medical images for training and validating neural networks. Methods Using our online collaborative platform 9 expert echocardiographers labelled 2056 images that comprised the training dataset. They labelled the four points from where the standard parasternal long axis (PLAX) measurements (interventricular septum, posterior wall, left ventricular dimension) would be made. Using these labelled images we trained a 2d convolutional neural network to replicate these labels. Separately, we curated an external validation dataset of the systolic and diastolic frames of 100 PLAX acquisitions. Each of these images were labelled twice by 13 different experts, and the average of the 26 measurements was taken as the consensus standard. We then compared the individual experts and the AI measurements on the external validation dataset to the consensus standard, and calculated the precision standard deviation (SD) of the signed differences from the consensus standard. Results For diastolic septum thickness, the AI had a precision SD of 1.8 mm (ICC 0.81; 95% CI 0.73 to 0.97), compared with 2.0 mm for the individual experts (ICC 0.64; 95% CI 0.57 to 0.72). For diastolic posterior wall thickness, the AI had a precision SD 1.4 mm (ICC 0.54; 95% CI 0.38 to 0.66), and the individual experts 2.2 mm (ICC 0.37; 95% CI 0.29 to 0.46). The AI's precision SD for left ventricular internal dimension was 3.5 mm (ICC 0.93, 95% CI 0.90 to 0.94), and for individual experts was 4.4mm (ICC 0.82, 95% CI 0.78 to 0.95). Both the experts and AI performed better in diastole than systole (precision SD AI 2.5mm vs 4.3mm, p&lt;0.0001; experts 3.3mm vs 5.3mm, p&lt;0.0001). Conclusions AI trained by a group of echocardiography experts was able to perform PLAX measurements which matched the reference standard more closely than any individual expert's own measurements. This open, collaborative approach may be a model for the development of AI that is explainable to, and trusted by clinicians. FUNDunding Acknowledgement Type of funding sources: Public grant(s) – National budget only. Main funding source(s): NIHR Imperil BRC ITMATDr Howard was additionally funded by Wellcome. Online collaborative platform Results of AI and experts

Assessing the Demand for Online Integration Platforms in the Real Estate Business

2020 ◽  
Vol 26 (1) ◽  
pp. 46-54
Author(s):  
N. S. Alekseeva ◽  
A. E. Druzhinin
Keyword(s):  
Real Estate ◽  
Real Estate Market ◽  
Passenger Transport ◽  
Integration Platform ◽  
Online Platform ◽  
The Real ◽  
The Real Estate ◽  
The Real Estate Market ◽  
Hospitality Sector ◽  
The Individual

Examination of the real estate market shows that implementing innovations in this field is a very difficult task. This prompts the question of the necessity of digitalizing the real estate business and of the demand for online integration platforms in this field.Aim. The presented study aims to assess the demand for online integration platforms in the real estate business.Tasks. The authors compare the online integration platforms in the real estate market with the online integration platforms in the hospitality sector and the individual passenger transport market in terms of the share of user profits that integrators receive for their services on the online platform; compare the share of user profits that integrators receive for their services on the online platform with the equivalent indicator in various other global economic activities; assess the value of the services provided by an online integration platform using the methodology proposed by G. G. Azgaldov and N. N. Karpova.Methods. The data were acquired from public Internet sources and personal interviews with the directors of companies that represent or employ the services of online integration platforms in St. Petersburg. The interviews and work with Internet sources were conducted in November 2019.Results. An integrator in the real estate market receives a share of profits of their service users that is comparable to that of the integrators in the hospitality sector and the individual passenger transport market. The share of profits of a Russian integrator in the real estate market is significantly higher than that in such industries as entertainment, fashion, or sports. The value of an online integration platform can be defined as highly valuable, since the expected value of this indicator is 1.5 times higher than the maximum table value.Conclusions. The performed analysis shows a high demand for online integration platforms in the real estate business. Market participants are willing to pay for the ability to use new digital technologies.

Sign in / Sign up

Export Citation Format

Share Document