Riccardo Munafò
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Riccardo Munafò has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
- Emiliano Votta
Politecnico di Milano
4 shared publications - Simone Saitta
Amsterdam University Medical Center
4 shared publications Eustachio Agricola
Unit of Cardiovascular Imaging, IRCCS San Raffaele Hospital, Milan, Italy; Vita-Salute San Raffaele University, Milan, Italy
3 shared publications- Giacomo Ingallina
Unit of Cardiovascular Imaging, IRCCS San Raffaele Hospital, Milan, Italy
3 shared publications - Davide Tondi
Department of Electronics, Information and Bioengineering, Politecnico di Milano, Milan, Italy
2 shared publications Giacomo Ingallina
2 shared publications
Latest Publications
- Open Access
- 27 September 2025
- Published by Elsevier in Computer Methods and Programs in Biomedicine
- Vol. 272, 109083
MitraClip device automated localization in 3D transesophageal echocardiography via deep learning
- 27 September 2025
- Published by Elsevier in Computer Methods and Programs in Biomedicine
- Vol. 272, 109083
Open AccessMitraClip device automated localization in 3D transesophageal echocardiography via deep learning
- 17 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-7
Vendor-Agnostic 3D Mitral Valve Segmentation Using Semi-Supervised Learning
Mitral valve assessment for disease diagnosis and treatment is commonly guided by ultrasound imaging, with 3D transesophageal echocardiography being the de facto standard modality. The complex 3D structure of the mitral valve poses a challenge for its time-efficient and accurate quantification. Deep neural networks have been proposed for automatic mitral valve segmentation, but variations in imaging devices and protocols across ultrasound vendors cause significant domain differences in echocardiography datasets, leading to suboptimal dataset-dependent segmentation techniques. In this work, we apply a semi-supervised learning method to develop a vendor-agnostic, automatic segmentation tool for 3D transesophageal echocardiography. We learn from one manually annotated ultrasound domain, to generate reliable pseudo-labels using an ensemble of three models, for another unseen ultrasound domain. Our teacher-student framework is validated on images from two of the biggest cardiac ultrasound manufacturers. In our experiments, the student model outperforms each teacher, with a Dice score of 82.20 ± 4.62% on known data and 71.32 ± 6.95% on unseen data, and an Average Surface Distance of 0.37 ± 0.07 mm and 0.82 ± 0.15 mm for the mitral valve on known and unknown domain respectively. These results enable efficient cross-domain analysis by reducing the need for manual annotation and ensuring consistent mitral valve analysis across different vendors.
- 17 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-7
Vendor-Agnostic 3D Mitral Valve Segmentation Using Semi-Supervised Learning
Mitral valve assessment for disease diagnosis and treatment is commonly guided by ultrasound imaging, with 3D transesophageal echocardiography being the de facto standard modality. The complex 3D structure of the mitral valve poses a challenge for its time-efficient and accurate quantification. Deep neural networks have been proposed for automatic mitral valve segmentation, but variations in imaging devices and protocols across ultrasound vendors cause significant domain differences in echocardiography datasets, leading to suboptimal dataset-dependent segmentation techniques. In this work, we apply a semi-supervised learning method to develop a vendor-agnostic, automatic segmentation tool for 3D transesophageal echocardiography. We learn from one manually annotated ultrasound domain, to generate reliable pseudo-labels using an ensemble of three models, for another unseen ultrasound domain. Our teacher-student framework is validated on images from two of the biggest cardiac ultrasound manufacturers. In our experiments, the student model outperforms each teacher, with a Dice score of 82.20 ± 4.62% on known data and 71.32 ± 6.95% on unseen data, and an Average Surface Distance of 0.37 ± 0.07 mm and 0.82 ± 0.15 mm for the mitral valve on known and unknown domain respectively. These results enable efficient cross-domain analysis by reducing the need for manual annotation and ensuring consistent mitral valve analysis across different vendors.
- Open Access
- 11 January 2025
- Published by Springer Nature in Medical & Biological Engineering & Computing
- Vol. 64 (8) , 1-16
Automatic 4D mitral valve segmentation from transesophageal echocardiography: a semi-supervised learning approach
Performing automatic and standardized 4D TEE segmentation and mitral valve analysis is challenging due to the limitations of echocardiography and the scarcity of manually annotated 4D images. This work proposes a semi-supervised training strategy using pseudo labelling for MV segmentation in 4D TEE; it employs a Teacher-Student framework to ensure reliable pseudo-label generation. 120 4D TEE recordings from 60 candidates for MV repair are used. The Teacher model, an ensemble of three convolutional neural networks, is trained on end-systole and end-diastole frames and is used to generate MV pseudo-segmentations on intermediate frames of the cardiac cycle. The pseudo-annotated frames augment the Student model’s training set, improving segmentation accuracy and temporal consistency. The Student outperforms individual Teachers, achieving a Dice score of 0.82, an average surface distance of 0.37 mm, and a 95% Hausdorff distance of 1.72 mm for MV leaflets. The Student model demonstrates reliable frame-by-frame MV segmentation, accurately capturing leaflet morphology and dynamics throughout the cardiac cycle, with a significant reduction in inference time compared to the ensemble. This approach greatly reduces manual annotation workload and ensures reliable, repeatable, and time-efficient MV analysis. Our method holds strong potential to enhance the precision and efficiency of MV diagnostics and treatment planning in clinical settings. Graphical
- 11 January 2025
- Published by Springer Nature in Medical & Biological Engineering & Computing
- Vol. 64 (8) , 1-16
Open AccessAutomatic 4D mitral valve segmentation from transesophageal echocardiography: a semi-supervised learning approach
Performing automatic and standardized 4D TEE segmentation and mitral valve analysis is challenging due to the limitations of echocardiography and the scarcity of manually annotated 4D images. This work proposes a semi-supervised training strategy using pseudo labelling for MV segmentation in 4D TEE; it employs a Teacher-Student framework to ensure reliable pseudo-label generation. 120 4D TEE recordings from 60 candidates for MV repair are used. The Teacher model, an ensemble of three convolutional neural networks, is trained on end-systole and end-diastole frames and is used to generate MV pseudo-segmentations on intermediate frames of the cardiac cycle. The pseudo-annotated frames augment the Student model’s training set, improving segmentation accuracy and temporal consistency. The Student outperforms individual Teachers, achieving a Dice score of 0.82, an average surface distance of 0.37 mm, and a 95% Hausdorff distance of 1.72 mm for MV leaflets. The Student model demonstrates reliable frame-by-frame MV segmentation, accurately capturing leaflet morphology and dynamics throughout the cardiac cycle, with a significant reduction in inference time compared to the ensemble. This approach greatly reduces manual annotation workload and ensures reliable, repeatable, and time-efficient MV analysis. Our method holds strong potential to enhance the precision and efficiency of MV diagnostics and treatment planning in clinical settings. Graphical
- Open Access
- 24 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Automatic 4D mitral valve segmentation from transesophageal echocardiography: a semi-supervised learning approach
Automatic and standardized methods for 4D transesophageal echocardiography (TEE) annotation and mitral valve (MV) motion and morphology analysis are difficult to achieve due to the known limitations of echocardiography and the limited availability of manually annotated 4D images. This work proposes a semi-supervised training strategy using pseudo-labelling to achieve MV segmentation in 4D TEE. The proposed strategy is based on a Teacher-Student framework that guarantees the generation of pseudo-labels with high reliability. 120 4D TEE recordings from 60 candidates to MV repair are used. The Teacher model, an ensemble of three convolutional neural networks (CNNs), is trained on end-systole (ES) and end-diastole (ES) frames and is used to generate MV pseudo segmentations on intermediate frames of the cardiac cycle. The pseudo annotated frames augment the training set of the Student model, enhancing its segmentation accuracy and temporal consistency. The Student model achieves superior performance than individual Teacher models with Dice Score of 0.82, average surface distance of 0.37 mm and 95% Hausdorff distance of 1.72 mm for the MV leaflets, and lower coordinate predictions error for the mitral annulus. The Student model demonstrates reliable frame-by-frame MV segmentation, accurately identifying leaflet morphology and dynamics throughout the cardiac cycle, with a significant reduction in inference time compared to the Ensemble model. This approach significantly reduces the manual annotation workload and ensures reliable, repeatable, and time-efficient MV analysis. Our method holds strong potential for improving the accuracy and efficiency of MV diagnostics and treatment planning in clinical settings.
- 24 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in TechRxiv
Open AccessAutomatic 4D mitral valve segmentation from transesophageal echocardiography: a semi-supervised learning approach
Automatic and standardized methods for 4D transesophageal echocardiography (TEE) annotation and mitral valve (MV) motion and morphology analysis are difficult to achieve due to the known limitations of echocardiography and the limited availability of manually annotated 4D images. This work proposes a semi-supervised training strategy using pseudo-labelling to achieve MV segmentation in 4D TEE. The proposed strategy is based on a Teacher-Student framework that guarantees the generation of pseudo-labels with high reliability. 120 4D TEE recordings from 60 candidates to MV repair are used. The Teacher model, an ensemble of three convolutional neural networks (CNNs), is trained on end-systole (ES) and end-diastole (ES) frames and is used to generate MV pseudo segmentations on intermediate frames of the cardiac cycle. The pseudo annotated frames augment the training set of the Student model, enhancing its segmentation accuracy and temporal consistency. The Student model achieves superior performance than individual Teacher models with Dice Score of 0.82, average surface distance of 0.37 mm and 95% Hausdorff distance of 1.72 mm for the MV leaflets, and lower coordinate predictions error for the mitral annulus. The Student model demonstrates reliable frame-by-frame MV segmentation, accurately identifying leaflet morphology and dynamics throughout the cardiac cycle, with a significant reduction in inference time compared to the Ensemble model. This approach significantly reduces the manual annotation workload and ensures reliable, repeatable, and time-efficient MV analysis. Our method holds strong potential for improving the accuracy and efficiency of MV diagnostics and treatment planning in clinical settings.
- Open Access
- 4 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 5295-5308
A Deep Learning-Based Fully Automated Pipeline for Regurgitant Mitral Valve Anatomy Analysis From 3D Echocardiography
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 time-consuming, 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 0.43 ± 0.14 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.
- 4 January 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 5295-5308
Open AccessA Deep Learning-Based Fully Automated Pipeline for Regurgitant Mitral Valve Anatomy Analysis From 3D Echocardiography
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 time-consuming, 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 0.43 ± 0.14 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.