Marie Schafstedde

Dr. Marie Schafstedde

orcid

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
30 September 2021 in IEEE Journal of Biomedical and Health Informatics
Verified
Journal Article
Closed access
PYPavlo Yevtushenko
Pavlo Yevtushenko
LGLeonid Goubergrits
Leonid Goubergrits
LGLina Gundelwein
Lina Gundelwein
ASArnaud Setio
Arnaud Setio
HRHeiko Ramm
Heiko Ramm
MSMarie Schafstedde
Marie Schafstedde
Image-based patient-specific modelling of hemodynamics are gaining increased popularity as a diagnosis and outcome prediction solution for a variety of cardiovascular diseases. While their potential to improve diagnostic capabilities and thereby clinical outcome is widely recognized, these methods require considerable computational resources since they are mostly based on conventional numerical methods such as computational fluid dynamics (CFD). As an alternative to the numerical methods, we propose a machine learning (ML) based approach to calculate patient-specific hemodynamic parameters. Compared to CFD based methods, our approach holds the benefit of being able to calculate a patient-specific hemodynamic outcome instantly with little need for computational power. In this proof-of-concept study, we present a deep artificial neural network (ANN) capable of computing hemodynamics for patients with aortic coarctation in a centerline aggregated (i.e., locally averaged) form. Considering the complex relation between vessels shape and hemodynamics on the one hand and the limited availability of suitable clinical data on the other, a sufficient accuracy of the ANN may however not be achieved with available data only. Another key aspect of this study is therefore the successful augmentation of available clinical data. Using a statistical shape model, additional training data was generated which substantially increased the ANN's accuracy, showcasing the ability of ML based methods to perform in-silico modelling tasks previously requiring resource intensive CFD simulations.

5 February 2021 in IEEE Transactions on Medical Imaging
Verified
Journal Article
Closed access
BTB. Thamsen
B. Thamsen
PYP. Yevtushenko
P. Yevtushenko
LGL. Gundelwein
L. Gundelwein
ASA. A. A. Setio
A. A. A. Setio
HLH. Lamecker
H. Lamecker
LGLeonid Goubergrits
Leonid Goubergrits
Modeling of hemodynamics and artificial intelligence have great potential to support clinical diagnosis and decision making. While hemodynamics modeling is extremely time- and resource-consuming, machine learning (ML) typically requires large training data that are often unavailable. The aim of this study was to develop and evaluate a novel methodology generating a large database of synthetic cases with characteristics similar to clinical cohorts of patients with coarctation of the aorta (CoA), a congenital heart disease associated with abnormal hemodynamics. Synthetic data allows use of ML approaches to investigate aortic morphometric pathology and its influence on hemodynamics. Magnetic resonance imaging data (154 patients as well as of healthy subjects) of aortic shape and flow were used to statistically characterize the clinical cohort. The methodology generating the synthetic cohort combined statistical shape modeling of aortic morphometry and aorta inlet flow fields and numerical flow simulations. Hierarchical clustering and non-linear regression analysis were successfully used to investigate the relationship between morphometry and hemodynamics and to demonstrate credibility of the synthetic cohort by comparison with a clinical cohort. A database of 2652 synthetic cases with realistic shape and hemodynamic properties was generated. Three shape clusters and respective differences in hemodynamics were identified. The novel model predicts the CoA pressure gradient with a root mean square error of 4.6 mmHg. In conclusion, synthetic data for anatomy and hemodynamics is a suitable means to address the lack of large datasets and provide a powerful basis for ML to gain new insights into cardiovascular diseases.

14 December 2021 in Circulation
Verified
Journal Article
Open access
PDF file
Nikolaus BerndtNikolaus Berndt
Nikolaus Berndt
JEJohannes Eckstein
Johannes Eckstein
IWIwona Wallach
Iwona Wallach
SNSarah Nordmeyer
Sarah Nordmeyer
MKMarcus Kelm
Marcus Kelm
Hermann-Georg HolzhütterHermann-Georg Holzhütter
Hermann-Georg Holzhütter
Background: Many heart diseases can result in reduced pumping capacity of the heart muscle. A mismatch between ATP demand and ATP production of cardiomyocytes is one of the possible causes. Assessment of the relation between myocardial ATP production (MV ATP ) and cardiac workload is important for better understanding disease development and choice of nutritional or pharmacologic treatment strategies. Because there is no method for measuring MV ATP in vivo, the use of physiology-based metabolic models in conjunction with protein abundance data is an attractive approach. METHOD: We developed a comprehensive kinetic model of cardiac energy metabolism (CARDIOKIN1) that recapitulates numerous experimental findings on cardiac metabolism obtained with isolated cardiomyocytes, perfused animal hearts, and in vivo studies with humans. We used the model to assess the energy status of the left ventricle of healthy participants and patients with aortic stenosis and mitral valve insufficiency. Maximal enzyme activities were individually scaled by means of protein abundances in left ventricle tissue samples. The energy status of the left ventricle was quantified by the ATP consumption at rest (MV ATP [rest]), at maximal workload (MV ATP [max]), and by the myocardial ATP production reserve, representing the span between MV ATP (rest) and MV ATP (max). Results: Compared with controls, in both groups of patients, MV ATP (rest) was increased and MV ATP (max) was decreased, resulting in a decreased myocardial ATP production reserve, although all patients had preserved ejection fraction. The variance of the energetic status was high, ranging from decreased to normal values. In both patient groups, the energetic status was tightly associated with mechanic energy demand. A decrease of MV ATP (max) was associated with a decrease of the cardiac output, indicating that cardiac functionality and energetic performance of the ventricle are closely coupled. Conclusions: Our analysis suggests that the ATP-producing capacity of the left ventricle of patients with valvular dysfunction is generally diminished and correlates positively with mechanical energy demand and cardiac output. However, large differences exist in the energetic state of the myocardium even in patients with similar clinical or image-based markers of hypertrophy and pump function. Registration: URL: https://www.clinicaltrials.gov ; Unique identifiers: NCT03172338 and NCT04068740.

1 July 2021 in European Journal of Cardio-Thoracic Surgery
Verified
Journal Article
Closed access
PDF file
MSMarie Schafstedde
Marie Schafstedde
SNSarah Nordmeyer
Sarah Nordmeyer
ASAnastasia Schleiger
Anastasia Schleiger
JNJohannes Nordmeyer
Johannes Nordmeyer
FBFelix Berger
Felix Berger
SOStanislav Ovroutski
Stanislav Ovroutski
The Fontan operation aims at reducing cyanosis and cardiac volume overload in patients with complex univentricular heart malformations. However, persisting or reoccurring cyanosis is frequently observed. We sought to systematically determine the prevalence and clinical consequences of persisting and secondary cyanosis after Fontan operation. A total of 331 Fontan patients, operated between 1984 and 2016 with a median postoperative follow-up of 7.9 (interquartile range 2.6–15.8) years, were studied retrospectively. Cyanosis was defined as transcutaneous oxygen saturation ≤93% at rest measured by pulse oximetry. Prevalence of cyanosis was analysed at 3 different time points (t1 = post-Fontan operation, t2 = post-Fontan cardiac catheterization, t3 = last follow-up) and the association of cyanosis with mortality was examined. Prevalence of cyanosis was 50% at t1 and 39% at t3. Fenestration was patent in 71% and 33% of all cyanotic patients at t1 and t3, respectively. In patients with clinical indication for catheterization (t2; n = 178/331), prevalence of cyanosis was 72%. At t2, patent fenestration (33%), veno-venous collaterals (24%) or both (32%) were present. Thirty-six (11%) patients died during follow-up. In a time-varying multivariable Cox regression analysis, cyanosis was the strongest predictor for late mortality (P < 0.001, hazard ratio 12.2, 95% confidence interval 3.7–40.5). Prevalence of cyanosis was considerable during long-term follow-up after Fontan operation and—as a surrogate parameter for unfavourable Fontan haemodynamics—is associated with increased late mortality. Accordingly, particular attention should be directed towards the persistence or reoccurrence of cyanosis during follow-up since it may indicate haemodynamic attrition and development of Fontan failure.

3 March 2023 in Frontiers in Cardiovascular Medicine
Verified
Journal Article
Open access
PYPavlo Yevtushenko
Pavlo Yevtushenko
LGLeonid Goubergrits
Leonid Goubergrits
BFBenedikt Franke
Benedikt Franke
TKTitus Kuehne
Titus Kuehne
MSMarie Schafstedde
Marie Schafstedde
Introduction: The computational modelling of blood flow is known to provide vital hemodynamic parameters for diagnosis and treatment-support for patients with valvular heart disease. However, most diagnosis/treatment-support solutions based on flow modelling proposed utilize time- and resource-intensive computational fluid dynamics (CFD) and are therefore difficult to implement into clinical practice. In contrast, deep learning (DL) algorithms provide results quickly with little need for computational power. Thus, modelling blood flow with DL instead of CFD may substantially enhances the usability of flow modelling-based diagnosis/treatment support in clinical routine. In this study, we propose a DL-based approach to compute pressure and wall-shear-stress (WSS) in the aorta and aortic valve of patients with aortic stenosis (AS).Methods: A total of 103 individual surface models of the aorta and aortic valve were constructed from computed tomography data of AS patients. Based on these surface models, a total of 267 patient-specific, steady-state CFD simulations of aortic flow under various flow rates were performed. Using this simulation data, an artificial neural network (ANN) was trained to compute spatially resolved pressure and WSS using a centerline-based representation. An unseen test subset of 23 cases was used to compare both methods.Results: ANN and CFD-based computations agreed well with a median relative difference between both methods of 6.0% for pressure and 4.9% for wall-shear-stress. Demonstrating the ability of DL to compute clinically relevant hemodynamic parameters for AS patients, this work presents a possible solution to facilitate the introduction of modelling-based treatment support into clinical practice.

8 September 2021 in Frontiers in Cardiovascular Medicine
Verified
Journal Article
Open access
BFBenedikt Franke
Benedikt Franke
JBJan Brüning
Jan Brüning
PYPavlo Yevtushenko
Pavlo Yevtushenko
HDHenryk Dreger
Henryk Dreger
ABAnna Brand
Anna Brand
MSMarie Schafstedde
Marie Schafstedde
Background: In patients with aortic stenosis, computed tomography (CT) provides important information about cardiovascular anatomy for treatment planning but is limited in determining relevant hemodynamic parameters such as the transvalvular pressure gradient (TPG).Purpose: In the present study, we aimed to validate a reduced-order model method for assessing TPG in aortic stenosis using CT data.Methods: TPGCT was calculated using a reduced-order model requiring the patient-specific peak-systolic aortic flow rate (Q) and the aortic valve area (AVA). AVA was determined by segmentation of the aortic valve leaflets, whereas Q was quantified based on volumetric assessment of the left ventricle. For validation, invasively measured TPGcatheter was calculated from pressure measurements in the left ventricle and the ascending aorta. Altogether, 84 data sets of patients with aortic stenosis were used to compare TPGCT against TPGcatheter.Results: TPGcatheter and TPGCT were 50.6 ± 28.0 and 48.0 ± 26 mmHg, respectively (p = 0.56). A Bland–Altman analysis revealed good agreement between both methods with a mean difference in TPG of 2.6 mmHg and a standard deviation of 19.3 mmHg. Both methods showed good correlation with r = 0.72 (p < 0.001).Conclusions: The presented CT-based method allows assessment of TPG in patients with aortic stenosis, extending the current capabilities of cardiac CT for diagnosis and treatment planning.