A method of parameter estimation for cardiovascular hemodynamics based on deep learning and its application to personalize a reduced‐order model

Zhou, Yang; He, Yuan; Wu, Jianwei; Cui, Chang; Chen, Minglong; Sun, Beibei

doi:10.1002/cnm.3533

Cited by 11 publications

(9 citation statements)

References 56 publications

(155 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When model complexity is too high for personalization, a reduced-order model can provide a useful alternative. For example, models with lower dimensionality can be developed 33,34 , or models can be reparameterized to reduce their number of parameters 35 . During our iterations of parameter subset reduction, we simplified the CircAdapt MultiPatch model by grouping regional LV parameters into global LV parameters, thereby simulating homogeneity of all LV wall segments.…”

Section: Discussionmentioning

confidence: 99%

Parameter Subset Reduction for Imaging-Based Digital Twin Generation of Patients with Left Ventricular Mechanical Discoordination

Koopsen,

Osta,

Loon

et al. 2023

Preprint

View full text Add to dashboard Cite

Background Integration of a patient’s non-invasive imaging data in a Digital Twin (DT) of the heart can provide valuable insight into the myocardial disease substrates underlying left ventricular (LV) mechanical discoordination. However, when generating a DT, model parameters should be identifiable to obtain robust parameter estimations. In this study, we used the CircAdapt model of the human heart and circulation to find a subset of parameters which were identifiable from LV cavity volume and regional strain measurements of patients with different substrates of left bundle branch block (LBBB) and myocardial infarction (MI). To this end, we included seven HF patients with LBBB (study ID: 2018 − 0863, registration date: 2019-10-07), of which four were non-ischemic (LBBB-only) and three had previous MI (LBBB-MI), and six narrow QRS patients with MI (MI-only) (study ID: NL45241.041.13, registration date: 2013-11-12). Morris Screening Method (MSM) was applied first to find parameters which were important for LV volume, regional strain, and strain rate indices. Second, this parameter subset was iteratively reduced based on parameter identifiability and reproducibility. Parameter identifiability was based on the diaphony calculated from quasi-Monte Carlo simulations and reproducibility was based on the intraclass correlation coefficient (\(ICC\)) obtained from repeated parameter estimation using dynamic multi-swarm particle swarm optimization. Goodness-of-fit was defined as the mean squared error (\({\chi }^{2}\)) of LV myocardial strain, strain rate, and cavity volume. Results A subset of 270 parameters remained after MSM which produced high-quality DTs of all patients (\({\chi }^{2}\)<1.6), but minimum parameter reproducibility was poor (\({ICC}_{min}\)=0.01). Iterative reduction yielded a reproducible (\({ICC}_{min}\)=0.83) subset of 75 parameters, including cardiac output, global LV activation duration, regional mechanical activation delay, and regional LV myocardial constitutive properties. This reduced subset produced patient-resembling DTs (\({\chi }^{2}\)<2.2), while septal-to-lateral wall workload imbalance was higher for the LBBB-only DTs than for the MI-only DTs (p < 0.05). Conclusions By applying sensitivity and identifiability analysis, we successfully determined a parameter subset of the CircAdapt model which can be used to generate imaging-based DTs of patients with LV mechanical discoordination. Parameters were reproducibly estimated using particle swarm optimization, and derived LV myocardial work distribution was representative for the patient’s underlying disease substrate. This DT technology enables patient-specific substrate characterization and can potentially be used to support clinical decision making.

show abstract

Section: Discussionmentioning

confidence: 99%

Parameter Subset Reduction for Imaging-Based Digital Twin Generation of Patients with Left Ventricular Mechanical Discoordination

Koopsen,

Osta,

Loon

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…For instance, CNN could be used to reconstruct the high−resolution turbulent field without solving the governing equation. The input of the network was the low−resolution flow field pooled from high−resolution flow field images obtained from the direct numerical simulation method (DNS), and the method was validated in laminar cylindrical flow and isotropic turbulence [ 147 ]. Similarly, CNN also could be used to perform the parameter estimation in cardiovascular hemodynamics [ 128 ].…”

Section: Application Of Artificial Intelligence In the Prediction Of ...mentioning

confidence: 99%

Interplay between Artificial Intelligence and Biomechanics Modeling in the Cardiovascular Disease Prediction

Liu

Deng

et al. 2022

Biomedicines

View full text Add to dashboard Cite

Cardiovascular disease (CVD) is the most common cause of morbidity and mortality worldwide, and early accurate diagnosis is the key point for improving and optimizing the prognosis of CVD. Recent progress in artificial intelligence (AI), especially machine learning (ML) technology, makes it possible to predict CVD. In this review, we first briefly introduced the overview development of artificial intelligence. Then we summarized some ML applications in cardiovascular diseases, including ML−based models to directly predict CVD based on risk factors or medical imaging findings and the ML−based hemodynamics with vascular geometries, equations, and methods for indirect assessment of CVD. We also discussed case studies where ML could be used as the surrogate for computational fluid dynamics in data−driven models and physics−driven models. ML models could be a surrogate for computational fluid dynamics, accelerate the process of disease prediction, and reduce manual intervention. Lastly, we briefly summarized the research difficulties and prospected the future development of AI technology in cardiovascular diseases.

show abstract

“…Experimentally, the model accuracy is sensitive to different waveform shapes 6 . Despite the fact that more functions and more parameters can result in higher accuracy, it is difficult to compare the parameters between methods, let alone with clinical indices 9–11 . Theoretically, model parameters are adaptively determined based on waveform morphologies, therefore it is necessary to further investigate their hemodynamic mechanisms to link clinical indices.…”

Section: Introductionmentioning

confidence: 99%

Interpretable and accurate curve‐fitting method for arterial pulse wave modeling and decomposition

Chen,

Peng,

Zhou

et al. 2023

Numer Methods Biomed Eng

View full text Add to dashboard Cite

Arterial pulse waveforms contain a wealth of information about the cardiovascular system. There is a lack of physical meaning in the mathematical model of arterial pulse waves, while the physical model fails to offer individuality as too many assumptions are involved. In this article, we focus on promoting the interpretability of the arterial pulse mathematical model. The proposed method is based on newly developed 3‐term fitting functions individualized by the physiological parameter assignment, which are the peak times of the reflected and dicrotic waves in a pulse. In this manner, the model allows decomposition of the pulse into sub‐signals with clear physiological significance. With nearly 10,000 pulse fitting experiments, it is demonstrated that the proposed method outperforms the standard methods in fitting accuracy while providing parameters linked to hemodynamic characteristics and common clinical indices such as the peripheral augmentation index (pAI). The proposed method innovatively maintains the individuality and accuracy of mathematical models while improving the interpretability of their parameters. The applications of this newly developed method, which explicitly incorporates hemodynamic characteristics, are expected to be particularly valuable in future pulse wave decomposition studies.

show abstract

A method of parameter estimation for cardiovascular hemodynamics based on deep learning and its application to personalize a reduced‐order model

Cited by 11 publications

References 56 publications

Parameter Subset Reduction for Imaging-Based Digital Twin Generation of Patients with Left Ventricular Mechanical Discoordination

Parameter Subset Reduction for Imaging-Based Digital Twin Generation of Patients with Left Ventricular Mechanical Discoordination

Interplay between Artificial Intelligence and Biomechanics Modeling in the Cardiovascular Disease Prediction

Interpretable and accurate curve‐fitting method for arterial pulse wave modeling and decomposition

Contact Info

Product

Resources

About