Skip to main content
European Commission logo print header

Quantitative characterization of cardiac tissue microstructure from Diffusion Tensor Imaging

Periodic Reporting for period 1 - DTI4micro (Quantitative characterization of cardiac tissue microstructure from Diffusion Tensor Imaging)

Reporting period: 2015-09-01 to 2017-08-31

Cardiac ventricular remodelling is the alteration of tissue microstructure and cardiac muscle structure and is a key factor for several cardiac diseases having profound effects on cardiac function. Nowadays, Cardiac Magnetic Resonance (CMR) is the gold-standard technique to assess in-vivo the heart structure and microstructure. This project has investigated advanced analytical tools to quantify and evaluate the heart structure and microstructure from CMR data and to investigate heart function in modern in-silico biophysical electrophysiological (EP) models. More specifically, we investigated how to characterize and incorporate quantitative characteristics of subject-specific CMR records in a cardiac function for EP simulations.
CMR techniques provide information in a non-invasive manner including multiple structural, tissue microstructure, and functional parameters. In my research, we focused on the use of standard CMR protocols for the subject-specific characterization of structure and microstructure of the heart. Additionally, we have developed a methodology for reconstructing the anatomy of the whole torso allowing the simulation of the electrocardiogram by placing virtual electrodes on the torso, and for comparing simulation results to recordings of the electrographic (ECG) signal.

Simulations of virtual hearts present a framework for the interpretation of medical data, allowing the assessment of biological hypothesis. In addition, they can be used to predict outcomes under simulated conditions such as cardiac or physiological dysfunctions, remodeling of the cardiac tissue, or drug treatments. The methods developed here will allow patient-specific structural, microstructural, and functional information of the heart from standard scans to be introduced in model personalization, providing a key step in their eventual application to routine clinical practice.

Summarizing, the achieved scientific objectives were:
- To develop techniques to accurately quantify macro- and microstructure of the heart tissue from CMR data valid for single subjects as well as for statistical analysis of populations.
- To develop a computational framework to personalize EP cardiac models from CMR data at the macro- and microstructural levels and to build personalized anatomies of the body surface allowing the evaluation of the EP simulation by direct comparison with ECG recordings.
- To investigate the use of personalized EP models in control and disease conditions, with a particular focus on macro- and microstructural cardiac diseases such as hypertrophic cardiomyopathy (HCM).
"The specific objectives of this project were:
- Reconstruction of the 3D heart anatomy from CMR recordings. Starting from clinical CMR recordings, we have developed an automatic pipeline for the reconstruction of the heart anatomy. Several techniques and methodologies have been employed in this complex task.
- Evaluation of microstructural tissue properties from CMR. Diffusion Tensor Imaging (DTI) allowed as to characterize the muscle microstructure measuring, among others, the orientation of the cardiac fibers and the degree of arrangement of the myocytes.
- Atlas-based reconstruction of a personalized torso. We tackle the challenging problem of reconstructing the body surface from standard clinical MRI scans with the use of statistical shape models.
- Simulation of patient-specific EP and their assessment on ECG end ECM recordings. Electrophysiological simulations were performed on subject-specific personalized models including anatomical, microstructural, and functional information from CMR data. We evaluated, on these virtual models, how the different physiological and structural parameters affect ECG signatures.


Overview of the results and their exploitation and dissemination.
- Dutta, Minchole, Zacur, Quinnc, Taggartd, Rodriguez, ""Early afterdepolarizations promote transmural reentry in ischemic human ventricles with reduced repolarization reserve"", Progress in Biophysics and Molecular Biology, 120:1-3, 2016.
- Villard, Zacur, Dall'Armellina, Grau, ""Correction of slice misalignment in multi-breath-hold cardiac MRI scans"", MICCAI Workshop on Statistical Atlases and Computational Models of the Heart (STACOM), Athens (Greece), Oct 2016.
- Villard, Carapella, Ariga, Grau, Zacur, ""Cardiac mesh reconstruction from sparse, heterogeneous contours"", Medical Image Understanding and Analysis (MIUA), Edinburgh (Scotland), Jul 2017.
- Zacur, Minchole, Villard, Carapella, Ariga, Rodriguez, Grau, ""MRI-based heart and torso personalization for computer modeling and simulation of cardiac electrophysiology"", MICCAI Workshop on Bio-Imaging and Visualization for Patient-Customized Simulations (BIVPCS), Quebec (Canada), Sep 2017.
- Villard, Grau, Zacur, ""Surface Mesh Reconstruction from Cardiac MRI Contours"", accepted in Journal of Imaging.
- Zacur, Grau, ""Parametric interpolation of PCA by geodesic regression"", submitted to IEEE Transactions on Pattern Analysis and Machine Intelligence.
- Zacur, Minchole, Villard, Carapella, Ariga, Rodriguez, Grau, ""MRI-based heart and torso personalization"", prepared to be submitted to Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization.
- Minchole, Zacur, Grau, Rodriguez, ""Dependency of the QRS complex with the MRI-based heart/torso geometries in computer models"", in preparation.
- Levrero-Florencio, Zacur, Minchole, Aguado-Sierra, Vazquez, Rodriguez, ""Patient-specific modelling of ventricular electro-mechanics: mathematical background, parameter estimation and image-based validation"", prepared to be submitted to Frontiers in Physiology.

Invited talks:
- British Heart Foundation Centre of Research Excellence annual symposium, September 2016.
- First international conference on Big Data on Biomedical Sciences, Xi'an, China, July 2017.
- Oxford imaging festival, Oxford, September 2017.

Exploitation and dissemination of results from secondary collaborations.
- Varela, Bisbal, Zacur, Berruezo, Aslandi, Mont, Lamata, ""Novel computational analysis of left atrial anatomy improves prediction of atrial fibrillation recurrence after ablation"", Frontiers in Physiology, 8:68, 2017.
- Warriner, Jackson, Zacur, Sammut, Sheridan, Hose, Lawford, Razavi, Niederer, Rinaldi, Lamata, ""An asymmetric wall-thickening pattern predicts response to Cardiac Resynchronization Therapy"", submitted to JACC: Cardiovascular Imaging.
"
Within the Institute of Biomedical Engineering and the Department of Engineering at the University of Oxford, I have participated in academic life in a number of ways, including the organization of some events.
Specifically, I have co-supervised two Master Thesis projects, co-supervised one project of the Centre for Doctoral Training in Biomedical Imaging. Also, I am currently co-supervising two Ph.D. candidates.

Besides the previously listed works, along the years enjoining the fellowship, I was able to start a solid collaborative network with several partners, such as King's College in London, the Oxford Centre for Clinical Magnetic Resonance Research at John Radcliffe Hospital in Oxford, the Department of Computer Science at the University of Oxford, Simula Research Laboratory in Norway and the Barcelona Supercomputing Center in Spain.
Volumetric model for personalized electrophysiological simulations.
Automatic imaging pipeline.
Epicardial and body surface potential results from personalized electrophysiological simulations.
Patient-specific reconstruction of heart anatomy.
Patient-specific torso reconstruction from CMR data.