SSSimone Saitta

Simone Saitta

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Simone Saitta has a total of 104 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Real‐World External Validation of Artificial Intelligence‐Based Full‐Vessel Segmentation for Intracoronary Optical Coherence Tomography

    by Rick H.J.A. Volleberg, Doosup Shin, Ruben G.A. van der Waerden, Craig R. Porter, Susan V. Thomas, Fernando Sosa, Simone Saitta, Aysun Cetinyurek‐Yavuz, Joske van der Zande, Thijs Luttikholt, et al.

    Background: Artificial intelligence (AI) allows automated evaluation of intracoronary optical coherence tomography (OCT) images. However, algorithms are mostly developed and validated on well‐curated data sets, which may not represent real‐world data. We sought to externally validate a previously developed algorithm performing full‐vessel segmentation for OCT in an unselected consecutive real‐world data set. Methods: This was a retrospective, single‐center, external validation study comprising 100 consecutive patients undergoing clinically indicated OCT. A previously developed AI algorithm (OCT‐AID) was used for automated pixelwise labeling of OCT images, distinguishing among lumen, guidewire artifact, intima, media, lipid plaque, calcium plaque, side branch, plaque rupture, thrombus, microvessel, and background. The AI‐based predictions were compared on a frame level to the reference standard obtained through manual OCT image analysis by expert readers. Results: Among 2560 analyzable frames, the agreement between the automated OCT image analysis and the reference standard was excellent for calcified plaque identification (κ=0.88 [95% CI, 0.84–0.92]) and quantification (intraclass correlation coefficient values ranged between 0.79 and 0.93), with a performance close to interobserver variability. For lipid plaque identification and quantification, the model performance was reasonable (κ=0.68 [95% CI, 0.64–0.72]; intraclass correlation coefficient for lipid arc, 0.79 [95% CI, 0.76–0.81]; intraclass correlation coefficient for minimum fibrous cap thickness, 0.59 [95% CI, 0.55–0.63]) and largely superior to interobserver variability. The algorithm performance for low‐prevalence features (e.g., plaque rupture) was limited. Conclusions: AI‐based fully automated evaluation of OCT images is feasible with performances consistent with interobserver variability in a real‐world data set of consecutive patients, supporting generalizability of the proposed methodology.


  • Open Access

    Artificial Intelligence-Derived Volumetric High-Risk Plaque Features Are Predominantly Located Proximal to Maximal Luminal Stenosis

    by Leah Heil, Thijs J. Luttikholt, Joske L. van der Zande, Jan-Quinten Mol, Ruben G.A. van der Waerden, Pierandrea Cancian, Xiaojin Gu, Clara I. Sánchez, Bram van Ginneken, Ivana Išgum, et al.


  • Open Access

    Near real-time multi-class segmentation for intravascular optical coherence tomography using knowledge distillation

    by Ruben G A van der Waerden, Rick H J A Volleberg, Pierandrea Cancian, Joske L van der Zande, Thijs J Luttikholt, Xiaojin Gu, Leah Heil, Jan-Quinten Mol, Kensuke Nishimiya, Tomasz Roleder, et al.

    Aims: Intravascular optical coherence tomography (OCT) enables high-resolution imaging of the coronary vessel wall, but manual image interpretation is time-consuming and existing automated approaches often require high computational resources and exhibit slow inference times, limiting clinical use. We developed OCT-AID-lite, a neural network for near-real-time multi-class OCT segmentation leveraging knowledge distillation and semi-supervised learning to accelerate inference while maintaining segmentation accuracy. Methods and results: A state-of-the-art model (OCT-AID) guided a compact U-Net–based student model (OCT-AID-lite) through knowledge distillation-based supervision. OCT-AID-lite was trained on 3466 manually annotated and 137 961 pseudo-labelled frames after automated quality control. On 389 internal test frames, OCT-AID-lite achieved a forward-pass time of 0.10 s for a 540-frame pullback, compared with 24.22 s for the OCT-AID model (P < 0.01). Including pre- and post-processing, total processing time was 5.50 s for OCT-AID-lite, compared with 30.62 s for OCT-AID (P < 0.01). For lipid and calcium plaque classification, OCT-AID-lite reached sensitivity/specificity of 98.1%/74.4% and 89.5%/87.5%, respectively. Pixel-wise segmentation performance on true-positive frames was high for guidewire, catheter, lumen, intima, and media (Dice: 0.79–0.99), moderate to high for sidebranch, lipid, and calcium (Dice: 0.76–0.78), and more variable for rare complex classes (Dice: 0.39–0.67). On an independent external test set, model predictions were in agreement with expert assessment. Conclusion: OCT-AID-lite enables accurate OCT segmentation in near real-time, allowing efficient quantitative characterization of plaque and vessel structures.


  • 4D Flow MRI ‐Based Quantification of Tricuspid Regurgitation: A Possible Approach to Improve Measurement Repeatability

    by Alessandra Riva, Giulia Pasqualin, Francesco Sturla, Lara Tondi, Simone Saitta, Davide Tondi, Jessica Zannoni, Giandomenico Disabato, Daniel Giese, Massimo Chessa, et al.

    Purpose: To address limitations in tricuspid regurgitation (TR) quantification by magnetic resonance imaging (MRI) or transthoracic echocardiography (TTE), a new method is proposed to directly quantify TR volume (TR V ) using 4D Flow MRI with retrospective tricuspid valve (TV) annular tracking. Methods: Thirty patients with TR were prospectively enrolled. TR V was quantified using four MRI‐based methods: (i) indirect, (ii) direct (2D phase‐contrast), (iii) 4D Flow with a static TV plane, and (iv) 4D Flow with dynamic annular tracking. Reproducibility and agreement were assessed in terms of TR grading against standard MRI‐ and TTE‐based methods. Results: The dynamic 4D Flow method correlated strongly with all other MRI methods ( r = 0.77–0.81, p < 0.0001) but consistently yielded higher TR V values. Compared with the indirect method, it showed a Bland–Altman bias of 16 mL (limits of agreement: −19 to 51 mL), along with the highest agreement (intraclass correlation coefficient of 0.88). Agreement in TR severity grading was poor across methods (kappa ≤ 0.07), with dynamic 4D Flow classifying more cases as moderate or severe. Agreement with TTE was slight (kappa = 0.17, p = 0.27) with concordant TR grading in 50% of patients. Inter‐observer reproducibility was excellent ( r = 0.99, bias = 0.7 mL, kappa = 0.80). Conclusion: 4D Flow with dynamic TV tracking offers a reproducible and potentially more sensitive method for TR V quantification than conventional MRI techniques. Despite limited alignment with TTE, this technique may help overcome key imaging limitations and improve TR clinical evaluation, especially in complex cases.

  • Building digital twins for personalized cardiovascular medicine: Advances, challenges, and future directions

    by Claudio Chiastra, Selene Pirola, Simone Saitta, Francesco Sturla, John F. LaDisa


  • Open Access

    MitraClip device automated localization in 3D transesophageal echocardiography via deep learning

    by Riccardo Munafò, Simone Saitta, Luca Vicentini, Davide Tondi, Veronica Ruozzi, Francesco Sturla, Giacomo Ingallina, Andrea Guidotti, Eustachio Agricola, Emiliano Votta


  • Open Access

    Artificial intelligence-based identification of thin-cap fibroatheromas and clinical outcomes: the PECTUS-AI study

    by Rick H J A Volleberg, Thijs J Luttikholt, Ruben G A van der Waerden, Pierandrea Cancian, Joske L van der Zande, Xiaojin Gu, Jan-Quinten Mol, Tomasz Roleder, Mathias Prokop, Clara I Sánchez, et al.

    Background and Aims: Coronary thin-cap fibroatheromas (TCFA) are associated with adverse outcome, but identification of TCFA requires expertise and is highly time-demanding. This study evaluated the utility of artificial intelligence (AI) for TCFA identification in relation to clinical outcome. Methods: The PECTUS-AI study is a secondary analysis from the prospective observational PECTUS-obs study, in which 438 patients with myocardial infarction underwent optical coherence tomography (OCT) of all fractional flow reserve-negative non-culprit lesions (i.e. target lesions). OCT images were analyzed for the presence of TCFA by an independent core laboratory (CL-TCFA) and OCT-AID, a recently developed and validated AI segmentation algorithm (AI-TCFA). The primary outcome was defined as the composite of death from any cause, non-fatal myocardial infarction or unplanned revascularisation at 2 years (±30 days), excluding procedural and stent-related events. Results: Among 414 patients, AI-TCFA and CL-TCFA were identified in 143 (34.5%) and 124 (30.0%) patients, respectively. AI-TCFA within the target lesion was significantly associated with the primary outcome [hazard ratio (HR) 1.99, 95% confidence interval (CI) 1.02–3.90, P = .04], while the HR for CL-TCFA was non-significant (1.67, 95% CI: .84–3.30, P = .14). When evaluating the complete pullback, AI-TCFA showed an even stronger association with the primary outcome (HR 5.50, 95% CI: 1.94–15.62, P < .001; negative predictive value 97.6%, 95% CI: 94.0%–99.3%). Conclusions: AI-based OCT image analysis allows standardized identification of patients at increased risk of adverse cardiovascular outcome, offering an alternative to manual image analysis. Furthermore, AI-assisted evaluation of complete imaged segments results in better prognostic discrimatory value than evaluation of the target lesion only.


  • Open Access

    World of Forms: Deformable geometric templates for one-shot surface meshing in coronary CT angiography

    by Rudolf L.M. van Herten, Ioannis Lagogiannis, Jelmer M. Wolterink, Steffen Bruns, Eva R. Meulendijks, Damini Dey, Joris R. de Groot, José P. Henriques, R. Nils Planken, Simone Saitta, et al.


  • Vendor-Agnostic 3D Mitral Valve Segmentation Using Semi-Supervised Learning

    by Maïlys Hau, Riccardo Munafò, Simone Saitta, Federico Veronesi, Giacomo Ingallina, Bjørnar Grenne, Frank Lindseth, Emiliano Votta, Gabriel Kiss

    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

    Comprehensive full-vessel segmentation and volumetric plaque quantification for intracoronary optical coherence tomography using deep learning

    by Rick H J A Volleberg, Ruben G A van der Waerden, Thijs J Luttikholt, Joske L van der Zande, Pierandrea Cancian, Xiaojin Gu, Jan-Quinten Mol, Silvan Quax, Mathias Prokop, Clara I Sánchez, et al.

    Aims: Intracoronary optical coherence tomography (OCT) provides detailed information on coronary lesions, but interpretation of OCT images is time-consuming and subject to interobserver variability. The aim of this study was to develop and validate a deep learning-based multiclass semantic segmentation algorithm for OCT (OCT-AID). Methods and results: A reference standard was obtained through manual multiclass annotation (guidewire artefact, lumen, side branch, intima, media, lipid plaque, calcified plaque, thrombus, plaque rupture, and background) of OCT images from a representative subset of pullbacks from the PECTUS-obs study. Pullbacks were randomly divided into a training and internal test set. An additional independent dataset was used for external testing. In total, 2808 frames were used for training and 218 for internal testing. The external test set comprised 392 frames. On the internal test set, the mean Dice score across nine classes was 0.659 overall and 0.757 on the true-positive frames, ranging from 0.281 to 0.989 per class. Substantial to almost perfect agreement was achieved for frame-wise identification of both lipid (κ=0.817, 95% CI 0.743–0.891) and calcified plaques (κ=0.795, 95% CI 0.703–0.887). For plaque quantification (e.g. lipid arc, calcium thickness), intraclass correlations of 0.664–0.884 were achieved. In the external test set, κ-values for lipid and calcified plaques were 0.720 (95% CI 0.640–0.800) and 0.851 (95% CI 0.794–0.908), respectively. Conclusion: The developed multiclass semantic segmentation method for intracoronary OCT images demonstrated promising capabilities for various classes, while having included difficult frames, such as those containing artefacts or destabilized plaques. This algorithm is an important step towards comprehensive and standardized OCT image interpretation.