Automatic detection of end‐diastolic and end‐systolic frames in 2D echocardiography

Zolgharni, Massoud; Negoita, Madalina; Dhutia, Niti M.; Mielewczik, Michael; Manoharan, K.; Sohaib, S.M. Afzal; Finegold, Judith A.; Sacchi, Stefano; Cole, Graham D.; Francis, Dárrel P.

doi:10.1111/echo.13587

Cited by 24 publications

(16 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this paper, we proposed a method to solve the ES and ED frame detection problem directly, without the need for LV segmentation. We reported results that were within interobserver variability error for detecting ED and ES frames (i.e., median disagreement of 3 frames [4]). We demonstrated the performance of several deep learning architectures, based on the combination of state-of-the-art CNN and RNN modules, and evaluated these architectures on a large dataset of echocardiography cine series.…”

Section: Discussionmentioning

confidence: 87%

See 1 more Smart Citation

Cardiac Phase Detection in Echocardiograms With Densely Gated Recurrent Neural Networks and Global Extrema Loss

Dezaki

Liao

Luong

et al. 2019

IEEE Trans. Med. Imaging

View full text Add to dashboard Cite

Accurate detection of end-systolic (ES) and end-diastolic (ED) frames in an echocardiographic cine series can be difficult but necessary pre-processing step for the development of automatic systems to measure cardiac parameters. The detection task is challenging due to variations in cardiac anatomy and heart rate often associated with pathological conditions. We formulate this problem as a regression problem and propose several deep learningbased architectures that minimize a novel global extrema structured loss function to localize the ED and ES frames. The proposed architectures integrate convolution neural networks (CNNs)-based image feature extraction model and recurrent neural networks (RNNs) to model temporal dependencies between each frame in a sequence. We explore two CNN architectures: DenseNet and ResNet, and four RNN architectures: long short-term memory, bi-directional LSTM, gated recurrent unit (GRU), and Bi-GRU, and compare the performance of these models. The optimal deep learning model consists of a DenseNet and GRU trained with the proposed loss function. On average, we achieved 0.20 and 1.43 frame mismatch for the ED and ES frames, respectively, which are within reported inter-observer variability for the manual detection of these frames.

show abstract

Section: Discussionmentioning

confidence: 87%

“…See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. Recently, Zolgharni et al [4] demonstrated that the median disagreement between five sonographers for the identification of ED and ES phases is 3 frames.…”

Section: Introductionmentioning

confidence: 99%

Cardiac Phase Detection in Echocardiograms With Densely Gated Recurrent Neural Networks and Global Extrema Loss

Dezaki

Liao

Luong

et al. 2019

IEEE Trans. Med. Imaging

View full text Add to dashboard Cite

show abstract

“…Their method allows analysis of potentially useful indices of left ventricle ejection/filling parameters. A recent study by Zolgharni [16] concluded that the speckle tracking by the development toolbox, which relies on the block matching algorithm is an efficient method for end‐diastole and end‐systole frames determination in 2D echocardiography sequences. One research [17] suggested three methods for estimation of the end‐diastole frame.…”

Section: Introductionmentioning

confidence: 99%

“…In the segmentation‐based approaches, end‐diastole and end‐systolic frames are characterised by the assumption that the most significant and smallest left ventricle segmented cross‐sections in a cardiac cycle correspond to the end‐systole and end‐diastole frames. In clinical practice, the end‐diastolic and end‐systolic frames are manually determined by using the R‐wave and T‐wave detection in the ECG signal, respectively [16]. The detection of the R‐peaks localisation in the noisy signal by using the Hilbert transform combined with a threshold technique was proposed in [19].…”

Section: Introductionmentioning

confidence: 99%

Novel approach for automatic mid‐diastole frame detection in 2D echocardiography sequences for performing planimetry of the mitral valve orifice

et al. 2020

View full text Add to dashboard Cite

“…Several models have received extensive attention in the ultrasound engineering community, such as block matching (BM) [20][21][22][23], optical flow [24][25][26], elastic registration [27,28], and machine learning models [29,30]. The most computationally efficient method of quantifying tissue motion on ultrasound is BM.…”

Section: Introductionmentioning

confidence: 99%

An optimisation-based iterative approach for speckle tracking echocardiography

Azarmehr

Howes

et al. 2020

Med Biol Eng Comput

Self Cite

View full text Add to dashboard Cite

Speckle tracking is the most prominent technique used to estimate the regional movement of the heart based on echocardiograms. In this study, we propose an optimised-based block matching algorithm to perform speckle tracking iteratively. The proposed technique was evaluated using a publicly available synthetic echocardiographic dataset with known ground-truth from several major vendors and for healthy/ischaemic cases. The results were compared with the results from the classic (standard) two-dimensional block matching. The proposed method presented an average displacement error of 0.57 pixels, while classic block matching provided an average error of 1.15 pixels. When estimating the segmental/regional longitudinal strain in healthy cases, the proposed method, with an average of 0.32 ± 0.53, outperformed the classic counterpart, with an average of 3.43 ± 2.84. A similar superior performance was observed in ischaemic cases. This method does not require any additional ad hoc filtering process. Therefore, it can potentially help to reduce the variability in the strain measurements caused by various post-processing techniques applied by different implementations of the speckle tracking.

show abstract

Automatic detection of end‐diastolic and end‐systolic frames in 2D echocardiography

Cited by 24 publications

References 24 publications

Cardiac Phase Detection in Echocardiograms With Densely Gated Recurrent Neural Networks and Global Extrema Loss

Cardiac Phase Detection in Echocardiograms With Densely Gated Recurrent Neural Networks and Global Extrema Loss

Novel approach for automatic mid‐diastole frame detection in 2D echocardiography sequences for performing planimetry of the mitral valve orifice

An optimisation-based iterative approach for speckle tracking echocardiography

Contact Info

Product

Resources

About