@conference {740, title = {ADC-Weighted Joint Registration-Estimation for Cardiac Diffusion Magnetic Resonance Imaging}, booktitle = {Congreso Anual de la Sociedad Espa{\~n}ola de Ingenier{\'\i}a Biom{\'e}dica}, year = {2017}, month = {12/2017}, abstract = {

The purpose of this work is to develop a method for the groupwise registration of diffusion weighted datasets of the heart which automatically provide smooth Apparent Diffusion Coefficient (ADC) estimations, by making use of a novel multimodal scheme. To this
end, we have introduced a joint methodology that simultaneously performs both the alignment of the images and the ADC estimation. In order to promote diffeomorphic transformations and to avoid undesirable noise amplification, we have included appropriate
smoothness constraints for both problems under the same formulation. The implemented multimodal registration metric incorporates the ADC estimation residuals, which are inversely weighted with the b-values to balance the influence of the signal level for each diffusion weighted image. Results show that the joint formulation provides more robust and precise ADC estimations and a significant improvement in the overlap of the contour
of manual delineations along the different b-values. The proposed algorithm is able to effectively deal with the presence of both physiological motion and inherent contrast variability for the different b-value images, increasing accuracy and robustness of the estimation of diffusion parameters for cardiac imaging.

}, author = {Santiago Sanz-Est{\'e}banez and J Royuela-del-Val and Jordi Broncano-Cabrero and Santiago Aja-Fern{\'a}ndez and Carlos Alberola-Lopez} } @conference {600, title = {An Automated Tensorial Classification Procedure for Left Ventricular Hypertrophic Cardiomyopathy}, booktitle = {IWBBIO 2016 (4th International Work-Conference on Bioinformatics and Biomedical Engineering)}, volume = {1}, year = {2016}, month = {2016}, pages = {1-12}, edition = {LNBI 9656}, address = {Granada, Spain}, abstract = {

Cardiovascular diseases are the leading cause of death globally. Therefore, classi cation tools play a major role in prevention and\ treatment of these diseases. Statistical learning theory applied to magnetic resonance imaging has led to the diagnosis of a variety of cardiomyopathies states.\ We propose a two-stage classi cation scheme capable of\ distinguishing between heterogeneous groups of hypertrophic cardiomyopathies and healthy patients.\ A multimodal processing pipeline is employed to estimate robust tensorial descriptors of myocardial mechanical\ properties for both short-axis and long-axis magnetic resonance tagged\ images using the least absolute deviation method. A homomorphic ltering procedure is used to align the cine segmentations to the tagged sequence and provides 3D tensor information in meaningful areas.\ Results\ have shown that the proposed pipeline provides tensorial measurements\ on which classi ers for the study of hypertrophic cardiomyopathies can\ be built with acceptable performance even for reduced samples sets.

}, keywords = {Fuzzy clustering, HARmonic Phase, Homomorphic Filtering, Hypertrophic Cardiomyopathy, Least Absolute Deviation, Magnetic Resonance Tagging, Support Vector Machines}, doi = {10.1007/978-3-319-31744-1 17}, author = {Santiago Sanz-Est{\'e}banez and J Royuela-del-Val and S. Merino-Caviedes and A. Revilla-Orodea and T. Sevilla-Ruiz and Martin-Fernandez, M and Carlos Alberola-Lopez} }