Systems and methods for determining a cardiovascular condition of a subject

1. A method for determining a cardiovascular condition of a subject, the method comprising:

capturing a three-dimensional electrocardiography image of the subject;

generating a two-dimensional cardiac map from the electrocardiography image; and

processing the cardiac map to determine the cardiovascular condition of the subject.

2. The method of claim 1 , wherein capturing a three-dimensional electrocardiography image of a subject comprises capturing a three-dimensional electrocardiography image of a subject using a smart shirt worn by the subject, the smart shirt comprising hundreds of electrodes adapted to measure electrical potentials around the subject's torso.

3. The method of claim 2 , wherein the smart shirt has a relaxed fit so that not every electrode is in contact with the subject's skin at all times when the shirt is worn.

4. The method of claim 3 , further comprising estimating missing data that results from electrodes losing contact with the subject's skin.

5. The method of claim 4 , wherein estimating missing data comprises generating a spatial model that models a spatial morphology of the electrocardiography image.

6. The method of claim 5 , wherein estimating missing data further comprises predicting a temporally evolving morphology of the electrocardiography image by recursively estimating time-varying parameters in the spatial model.

7. The method of claim 1 , wherein generating a two-dimensional cardiac map comprises normalizing the electrocardiography image onto a unit sphere.

8. The method of claim 7 , wherein generating a two-dimensional cardiac map further comprises projecting the normalized electrocardiography image onto a two-dimensional surface to create the cardiac map.

9. The method of claim 1 , wherein processing the cardiac maps comprises decomposing the cardiac map into a hierarchical modeling structure.

10. The method of claim 9 , wherein the hierarchical modeling structure comprises a global morphology, a local variability, and noise.

11. The method of claim 10 , wherein processing the cardiac maps further comprises estimating parameters in the hierarchical modeling structure.

12. The method of claim 11 , wherein processing the cardiac maps further comprises evaluating the cardiovascular condition of the subject with the global morphology and the local variability.

13. A non-transitory computer-readable medium that stores an electrocardiography data processing system comprising:

logic configured to receive a three-dimensional electrocardiography image of a subject;

logic configured to generate a two-dimensional cardiac map from the electrocardiography image; and

logic configured to process the cardiac map to determine the cardiovascular condition of the subject.

14. The computer-readable medium of claim 13 , further comprising logic configured to estimate missing data that results from electrodes losing contact with the subject's skin.

15. The computer-readable medium of claim 14 , wherein the logic configured to estimate missing data comprises logic configured to generate a spatial model that models a spatial morphology of the electrocardiography image and predict a temporally evolving morphology of the electrocardiography image by recursively estimating time-varying parameters in the spatial model.

16. The computer-readable medium of claim 13 , wherein the logic configured to generate a two-dimensional cardiac map is configured to normalize the electrocardiography image onto a unit sphere and project the normalized electrocardiography image onto a two-dimensional surface to create the cardiac map.

17. The computer-readable medium of claim 13 , wherein the logic configured to process the cardiac map is configured to decompose the cardiac map into a hierarchical modeling structure including a global morphology, a local variability, and noise, estimate parameters in the hierarchical modeling structure, and evaluate the cardiovascular condition of the subject with the global morphology and the local variability.