Cardiovascular imaging plays a crucial role in guiding diagnosis and supporting personalized treatments. To this purpose, cardiovascular imaging is also used to feed computer-based models, which in turn are increasingly exploited to investigate cardiovascular dysfunctions, support the clinical decision-making process, and plan tailored interventions.
However, despite a significant increase in computational capabilities and spreading of novel technologies able to automatize the workflow (e.g., AI-based image segmentation), translating personalized and computer-based solutions into the daily clinical practice is still a challenge. To achieve this, an effective trade-off between complexity, reliability – as measured against ground-truth in vivo data – and computational time of the modeling strategy is mandatory.
The fine granularity offered by computational analysis is frequently combined with medical imaging to advance cardiovascular research through personalized solutions for diagnosis and treatment of cardiovascular disease. Ground-truth imaging is exploited to feed patient-specific models providing anatomical realism and in vivo boundary/loading conditions to minimize model assumptions. At the same time, computational models should be verified and validated against patients’ data. However, in practice, several issues remain unsolved. Anatomically realistic cardiovascular models provide unique opportunity to predict the outcome associated with different procedures and medical devices, but lack of reliable boundary conditions can inevitably jeopardize both clinical relevance and applicability of the analysis. Additional quantitative data can be directly retrieved from in vivo imaging. Yet, this process can be non-trivial, often requiring additional imaging, and standardization of data extraction is still lacking. Furthermore, the computational workflow should be accomplished within a time-frame as compatible as possible with clinical practice. However, both the demanding hardware requirements and the high computational costs are recognized as key factors limiting the usability of accurate computational models in real clinical settings. Hence, searching for a trade-off between accuracy and computational cost still deserves further efforts.
The present Research Topic aims at sharing cutting-edge computational solutions in the field of cardiovascular research with a specific focus on improving their reliability and clinical usability through integration of cardiovascular imaging, measurement data, and minimization of the computational expense. Topics of interest include, but are not limited to:
• Novel image-based methods to enhance the reliability of patient-specific modeling
• User-friendly computational approaches to personalize the treatment of cardiovascular disease as well as the planning of cardiovascular interventions
• Hybrid computational-experimental approaches embedding emerging technologies (e.g., data enhancement, reduced order modeling, data assimilation, and artificial intelligence) to improve patient-specific analysis of cardiovascular disease and interventions
• Advanced methods to investigate clinically-relevant pre-operative risk factors and predict the feasibility and outcome of procedures
• Model verification, validation and uncertainty quantification (VVUQ), and optimization strategies to speed up patient-specific simulations with adequate trade-off between model accuracy and complexity
• Advances and novel proof-of-concept analyses in translating cardiovascular computational methods to the clinic.
Computational modeling should be at the core of the proposed studies. Both Original Research and Review articles on one or more of the themes above are welcome.
Cardiovascular imaging plays a crucial role in guiding diagnosis and supporting personalized treatments. To this purpose, cardiovascular imaging is also used to feed computer-based models, which in turn are increasingly exploited to investigate cardiovascular dysfunctions, support the clinical decision-making process, and plan tailored interventions.
However, despite a significant increase in computational capabilities and spreading of novel technologies able to automatize the workflow (e.g., AI-based image segmentation), translating personalized and computer-based solutions into the daily clinical practice is still a challenge. To achieve this, an effective trade-off between complexity, reliability – as measured against ground-truth in vivo data – and computational time of the modeling strategy is mandatory.
The fine granularity offered by computational analysis is frequently combined with medical imaging to advance cardiovascular research through personalized solutions for diagnosis and treatment of cardiovascular disease. Ground-truth imaging is exploited to feed patient-specific models providing anatomical realism and in vivo boundary/loading conditions to minimize model assumptions. At the same time, computational models should be verified and validated against patients’ data. However, in practice, several issues remain unsolved. Anatomically realistic cardiovascular models provide unique opportunity to predict the outcome associated with different procedures and medical devices, but lack of reliable boundary conditions can inevitably jeopardize both clinical relevance and applicability of the analysis. Additional quantitative data can be directly retrieved from in vivo imaging. Yet, this process can be non-trivial, often requiring additional imaging, and standardization of data extraction is still lacking. Furthermore, the computational workflow should be accomplished within a time-frame as compatible as possible with clinical practice. However, both the demanding hardware requirements and the high computational costs are recognized as key factors limiting the usability of accurate computational models in real clinical settings. Hence, searching for a trade-off between accuracy and computational cost still deserves further efforts.
The present Research Topic aims at sharing cutting-edge computational solutions in the field of cardiovascular research with a specific focus on improving their reliability and clinical usability through integration of cardiovascular imaging, measurement data, and minimization of the computational expense. Topics of interest include, but are not limited to:
• Novel image-based methods to enhance the reliability of patient-specific modeling
• User-friendly computational approaches to personalize the treatment of cardiovascular disease as well as the planning of cardiovascular interventions
• Hybrid computational-experimental approaches embedding emerging technologies (e.g., data enhancement, reduced order modeling, data assimilation, and artificial intelligence) to improve patient-specific analysis of cardiovascular disease and interventions
• Advanced methods to investigate clinically-relevant pre-operative risk factors and predict the feasibility and outcome of procedures
• Model verification, validation and uncertainty quantification (VVUQ), and optimization strategies to speed up patient-specific simulations with adequate trade-off between model accuracy and complexity
• Advances and novel proof-of-concept analyses in translating cardiovascular computational methods to the clinic.
Computational modeling should be at the core of the proposed studies. Both Original Research and Review articles on one or more of the themes above are welcome.