Integrated Multi-Modal Data Analysis for Computational Modeling of Healthy and MI Conditions in Porcine Hearts - Experimental Dataset

This repository contains the experimental dataset associated with the publication: “Integrated Multi-Modal Data Analysis for Computational Modeling of Healthy and Location-Dependent Myocardial Infarction Conditions in Porcine Hearts” by Ricardo M. Rosales, Ming Wu, Piet Claus, Stefan Janssens, Gonzalo R. Ríos-Muñoz, María Eugenia Fernández-Santos, Pablo Martinez-Legazpi, Javier Bermejo, Aiden Flanagan, Manuel Doblaré, Ana Mincholé, and Esther Pueyo. For the associated modeling and simulation dataset, please refer to the Simulation Dataset. Data Structure The dataset is organized into 15 individual folders, each corresponding to a specific porcine subject. For a comprehensive overview of the experimental data types collected per subject, please refer to Table 1 in the Preprint Paper. 1. Imaging Data (CMR) All cardiac magnetic resonance (CMR) data was processed using 3DSlicer. Cine CMR: Multi-volume data provided in .nrrd format. Each cine data includes a corresponding segmentation file (ending in .seg.nrrd) for the end-diastolic timepoint. Late Gadolinium Enhancement (LGE): Provided in .nii (NIfTI) format. Segmentations (.seg) identify healthy and affected zones. A "remote healthy region" is included to facilitate scar/border zone quantification via the full-width half-maximum method. Diffusion-Weighted (DW) MRI: Provided in .nrrd format with corresponding .seg segmentation files. Thoracic MRI: Specific to Pig 6, this volume shows ECG lead placement and is aligned with the subject's LGE data. The file MR_thorax_ecg_leads.mrk contains markups that can be loaded in Slicer to visualize electrode positioning. 2. Optical Mapping (OM) Available within specific subject folders under the OM/ directory: Videos: Raw optical mapping recordings in MATLAB (.mat) format. Metadata (info.xlsx): Detailed stimulation parameters, spatial resolution, sampling frequency, and specified Regions of Interest (ROIs). Stimulations types can be SR (sinus rhythm), RA (right atria), LV (left ventricle) or RV (right ventricle). ROIs: Defined regions for healthy, MI, and global cardiac analysis. Usage: the script getResultsFromOmapFiltered.py in simOmPP can be used after filtering the OM videos for obtaining the tissue-level electrophysiological outputs (such as activation times, conduction velocities and action potential durations) together with the ROIs included in this dataset 3. Electroanatomical Mapping (EAM) EAM data for Pig 6 is provided in .vtk format (Paraview can be used for visualization): EAM_aorta_anatomy_and_maps: Mesh data including unipolar/bipolar activation and voltage values for the aorta, used for aligning EAM points with the CMR-derived model of pig 6. EAM_lv_anatomy_and_maps: Mesh data including unipolar/bipolar activation and voltage values interpolated from the measurements in the Left Ventricle (LV). EAM_lv_anatomy_and_maps_points: Point data measurements including unipolar/bipolar activation and voltage values for the LV used to define post-MI subject-specific activation patterns. 4. Electrocardiogram (ECG) The repository includes 12-lead ECG median beats provided in MATLAB (.mat) format: Healthy Condition: Located in the Pig 15 folder. We thank Dylan Vermoortele from the Katholieke Universiteit Leuven for providing this data. MI Condition: Located in the Pig 6 folder (derived from EAM data). Usage Note: For further technical details regarding the processing pipelines, simulation parameters, or code, please refer to the primary paper and the linked software repositories.