A Bottom-Up Approach for Real-Time Mitral Valve Annulus Modeling on 3D Echo Images

Zhang, Yue; Amadou, Abdoul-aziz; Voigt, Ingmar; Mihalef, Viorel; Houle, Helene; John, Matthias; Mansi, Tommaso; Liao, Rui

doi:10.1007/978-3-030-59725-2_44

Cited by 4 publications

(16 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Zhang et al [31] report a surgical view prediction error that is significantly larger than our experiment. However, as they use three points to calculate each plane, their plane estimates are more susceptible to small coordinate prediction errors.…”

Section: Measurement Error [Unit]contrasting

confidence: 62%

“…Further, the model obtained a surgical view prediction error of 3.28 ± 2.92 • and a relative perimeter error of 5.8 ± 4.8%. Table 4 presents the coordinate prediction metrics introduced in Section III-D with the respective comparable results from [25], [29], [30] and [31] for comparison.…”

Section: Resultsmentioning

confidence: 99%

“…The same applies to several other tools in clinical use and the methods presented in [25] and [29]. Compared to the above methods, a benefit of our work is that [25], [29], [31], and [30]. An overview of the metrics is given in Section III-D.…”

Section: A Comparison With Existing Methodsmentioning

confidence: 99%

“…A limitation of [30] is that it did not use 3D context but obtained 3D predictions using an iterative post-processing algorithm along the 2D planes -where the prediction in a plane relied on the result of the neighboring plane. Zhang et al [31] presented another mitral annulus segmentation pipelineusing a combination of deep reinforcement learning, 2D landmark detection using a 2D CNN, and spline fitting to produce the 3D annulus predictions.…”

Section: A Related Workmentioning

confidence: 99%

“…However, as they use three points to calculate each plane, their plane estimates are more susceptible to small coordinate prediction errors. The main metric for mitral annulus predictions in [31] is the curve-to-curve metric. In our experiment, only three of 135 volumes (from examination 13) have a larger curve-to-curve error than 3.5 mm -the reported average prediction error in [31].…”

Section: Measurement Error [Unit]mentioning

confidence: 99%

See 4 more Smart Citations

Mitral Annulus Segmentation and Anatomical Orientation Detection in TEE Images Using Periodic 3D CNN

et al. 2022

View full text Add to dashboard Cite

Segmentation of the mitral annulus is often an important step in cardiac examinations. We propose a robust 3D method for predicting the anatomical orientation and segmentation of the mitral annulus in 3D transesophageal echocardiography. The method takes advantage of the circular anatomy of the annulus by utilizing cylinder coordinate samples and a 3D convolutional neural network with circular convolutions. Furthermore, the paper proposes new landmark detection loss functions based on the earth mover's distance. The method's effectiveness was demonstrated by training a HighRes3dNet model and evaluating its performance on a separate test set consisting of 135 frames from 19 examinations. The obtained coordinate prediction error was 1.96±1.62 mm, and the anatomical orientation prediction error was 9.7 • ±15.8 • . The robust and fully automatic mitral annulus segmentation and orientation prediction provided by the method can ease the workload of clinicians and provide time savings in clinics.

show abstract

Section: Measurement Error [Unit]contrasting

confidence: 62%

Section: Resultsmentioning

confidence: 99%

Section: A Comparison With Existing Methodsmentioning

confidence: 99%

Section: A Related Workmentioning

confidence: 99%

Section: Measurement Error [Unit]mentioning

confidence: 99%

See 3 more Smart Citations

Mitral Annulus Segmentation and Anatomical Orientation Detection in TEE Images Using Periodic 3D CNN

et al. 2022

View full text Add to dashboard Cite

show abstract

Introducing Anatomical Constraints in Mitral Annulus Segmentation in Transesophageal Echocardiography

Andreassen,

Thomas,

Solberg

et al. 2024

Lecture Notes in Computer Science

View full text Add to dashboard Cite

A Deep Learning-Based Fully Automated Pipeline for Regurgitant Mitral Valve Anatomy Analysis From 3D Echocardiography

Munafò,

Saitta,

Ingallina

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Three-dimensional transesophageal echocardiography (3DTEE) is the recommended imaging technique for the assessment of mitral valve (MV) morphology and lesions in case of mitral regurgitation (MR) requiring surgical or transcatheter repair. Such assessment is key to thorough intervention planning and to intraprocedural guidance. However, it requires segmentation from 3DTEE images, which is timeconsuming, operator-dependent, and often merely qualitative. In the present work, a novel workflow to quantify the patient-specific MV geometry from 3DTEE is proposed. The developed approach relies on a 3D multi-decoder residual convolutional neural network (CNN) with a U-Net architecture for multi-class segmentation of MV annulus and leaflets. The CNN was trained and tested on a dataset comprising 55 3DTEE examinations of MR-affected patients. After training, the CNN is embedded into a fully automatic, and hence fully repeatable, pipeline that refines the predicted segmentation, detects MV anatomical landmarks and quantifies MV morphology. The trained 3D CNN achieves an average Dice score of 0.82 ± 0.06, mean surface distance of 0.95 ± 0.27 mm and 95% Hausdorff Distance (HD) of 3.57 ± 1.56 mm before segmentation refinement, outperforming a state-of-the-art baseline residual U-Net architecture, and provides an unprecedented multi-class segmentation of the annulus, anterior and posterior leaflet. The automatic 3D linear morphological measurements of the annulus and leaflets, specifically diameters and lengths, exhibit differences of less than 1.45 mm when compared to ground truth values. These measurements also demonstrate strong overall agreement with analyses conducted by semi-automated commercial software. The whole process requires minimal user interaction and requires approximately 15 seconds.

show abstract

A Bottom-Up Approach for Real-Time Mitral Valve Annulus Modeling on 3D Echo Images

Cited by 4 publications

References 13 publications

Mitral Annulus Segmentation and Anatomical Orientation Detection in TEE Images Using Periodic 3D CNN

Mitral Annulus Segmentation and Anatomical Orientation Detection in TEE Images Using Periodic 3D CNN

Introducing Anatomical Constraints in Mitral Annulus Segmentation in Transesophageal Echocardiography

A Deep Learning-Based Fully Automated Pipeline for Regurgitant Mitral Valve Anatomy Analysis From 3D Echocardiography

Contact Info

Product

Resources

About