Calibration of simulated cardiograms
Systems are provided for generating data representing electromagnetic states of a heart for medical, scientific, research, and/or engineering purposes. The systems generate the data based on source configurations such as dimensions of, and scar or fibrosis or pro-arrhythmic substrate location within, a heart and a computational model of the electromagnetic output of the heart. The systems may dynamically generate the source configurations to provide representative source configurations that may be found in a population. For each source configuration of the electromagnetic source, the systems run a simulation of the functioning of the heart to generate modeled electromagnetic output (e.g., an electromagnetic mesh for each simulation step with a voltage at each point of the electromagnetic mesh) for that source configuration. The systems may generate a cardiogram for each source configuration from the modeled electromagnetic output of that source configuration for use in predicting the source location of an arrhythmia.
1. A method performed by one or more computing systems for identifying simulated electromagnetic data of an electromagnetic source that are similar to a patient electromagnetic data of a patient electromagnetic source, the patient electromagnetic data generated based on patient pacing locations of the patient electromagnetic source, the method comprising:
for each of a plurality of source configurations of the electromagnetic source, generating simulated electromagnetic output of the electromagnetic source by simulating, using a finite-element mesh, electrical activity of the electromagnetic source based on simulated pacing locations of the electromagnetic source;
generating a library of simulated electromagnetic data by, for at least some of the simulated electromagnetic output, deriving simulated electromagnetic data from the simulated electromagnetic output;
identifying, from the library, simulated electromagnetic data generated based on simulated pacing locations that are similar to the patient pacing locations;
identifying, from the library, simulated electromagnetic data generated based on simulated action potentials that are similar to patient action potentials;
identifying, from the library, simulated electromagnetic data generated based on simulated conduction velocities that are similar to patient conduction velocities;
identifying, from the library, simulated electromagnetic data generated based on a simulation configuration parameter that is similar to the patient configuration parameter of the patient electromagnetic source, the simulation configuration parameter and the patient configuration parameter being other than a pacing location, an action potential, and a conduction velocity; and
generating a calibrated collection of simulated electromagnetic data based on the identified simulated electromagnetic data.
2. The method of claim 1 further comprising identifying, from the library, simulated electromagnetic data that are similar to patient electromagnetic data.
3. The method of claim 1 wherein a pacing lead is placed with an invasive pacing device that administers electromagnetic pulses from inside the electromagnetic source.
4. The method of claim 3 wherein the invasive pacing device is a catheter, an implantable cardioverter, or a pacemaker.
5. The method of claim 1 wherein the patient pacing is performed using a non-invasive pacing device that administers electromagnetic pulses from outside the electromagnetic source.
6. The method of claim 5 wherein the non-invasive pacing device generates an electromagnetic field to pace the patient electromagnetic source.
7. The method of claim 6 wherein the non-invasive pacing device is a magnetic resonance device.
8. The method of claim 1 wherein the electromagnetic source is a heart, the simulated electromagnetic data is a cardiogram, and the patient electromagnetic data is a cardiogram.
9. The method of claim 8 wherein the cardiogram is a vectorcardiogram.
10. The method of claim 8 wherein the cardiogram is an electrocardiogram.
11. The method of claim 1 wherein the electromagnetic source is a brain.
12. The method of claim 1 wherein the configuration parameter is orientation.
13. The method of claim 1 wherein at least one of the identifying of simulated electromagnetic data is based on a set of previously identified simulated electromagnetic data.
14. The method of claim 1 further comprising identifying from the calibrated collection a simulated electromagnetic data that is similar to a patient electromagnetic data to identify a configuration parameter not collected from the patient.
15. The method of claim 14 wherein the configuration parameter is source location.
16. The method of claim 1 further comprising identifying simulated electromagnetic data that is similar to patient electromagnetic data by identifying cycles within the patient electromagnetic data and, for each of a plurality of simulated electromagnetic data of the calibrated collection, calculating a similarity score indicating similarity between the identified cycles and cycles of the simulated electromagnetic data.
17. A method performed by one or more computing systems for identifying simulated electromagnetic data of an electromagnetic source that are similar to a patient electromagnetic data of a patient electromagnetic source, the simulated electromagnetic data generated based on simulated pacing locations of the electromagnetic source and the patient electromagnetic data generated based on patient pacing locations of the patient electromagnetic source, the method comprising:
identifying simulated electromagnetic data generated based on simulated pacing locations that are similar to the patient pacing locations;
identifying simulated electromagnetic data generated based on simulated action potentials that are similar to patient action potentials;
identifying simulated electromagnetic data generated based on simulated conduction velocities that are similar to patient conduction velocities; and
generating training data that includes at least some of the identified simulated electromagnetic data labeled with a label that is a configuration parameter not collected from the patient; and
training a patient-specific model classifier based on the training data.
18. The method of claim 17 wherein the label is a source location.
19. The method of claim 17 further comprising applying the patient-specific model classifier to patient electromagnetic data to identify the configuration parameter that is not collected from the patient.
20. The method of claim 19 further comprising outputting the identified configuration parameter to inform treatment of the patient.
21. A method performed by one or more computing systems for identifying simulated electromagnetic data of an electromagnetic source that are similar to a patient electromagnetic data of a patient electromagnetic source, the simulated electromagnetic data generated based on simulated pacing locations of the electromagnetic source and the patient electromagnetic data generated based on patient pacing locations of the patient electromagnetic source, the method comprising:
identifying simulated electromagnetic data generated based on simulated pacing locations that are similar to the patient pacing locations;
identifying simulated electromagnetic data generated based on simulated action potentials that are similar to patient action potentials;
identifying simulated electromagnetic data generated based on simulated conduction velocities that are similar to patient conduction velocities; and
training a mapping function that maps the identified simulated electromagnetic data to the patient electromagnetic data; and
applying the mapping function to a set of simulated electromagnetic data to generate transformed simulated electromagnetic data wherein a calibrated collection of simulated electromagnetic data includes the transformed simulated electromagnetic data.
22. The method of claim 21 further comprising training a patient-specific classifier based on the calibrated collection for identifying a source location based on a patient electromagnetic data.
23. One or more computing systems for identifying simulated electromagnetic data of an electromagnetic source that are similar to a patient electromagnetic data of a patient electromagnetic source, the patient electromagnetic data generated based on a patient pacing locations of the patient electromagnetic source, the one or more computing systems comprising:
one or more computer-readable storage mediums for storing computer-executable instructions for controlling the one or more computing systems to:
access a library of simulated electromagnetic data, the library generated by deriving simulated electromagnetic data from simulated electromagnetic output of the electromagnetic source, the simulated electromagnetic output generated by, for each of a plurality of source configurations of the electromagnetic source, simulating electrical activity of the electromagnetic source using a finite-element mesh to generate the simulated electromagnetic output of the electromagnetic source;
identify, from the library, simulated electromagnetic data generated based on simulated pacing locations that are similar to the patient pacing locations;
identify, from the library, simulated electromagnetic data generated based on simulated action potentials that are similar to patient action potentials;
identify, from the library, simulated electromagnetic data generated based on simulated conduction velocities that are similar to patient conduction velocities;
identify, from the library, simulated electromagnetic data generated based on a simulation configuration parameter that is similar to a characteristic of the patient electromagnetic source corresponding to the simulation configuration parameter, the simulation configuration parameter being other than a pacing location, an action potential, and a conduction velocity; and
generate a calibrated collection of simulated electromagnetic data based on the identified simulated electromagnetic data; and
one or more processors for executing the computer-executable instructions stored in the one or more computer-readable storage mediums.
24. The one or more computing systems of claim 23 wherein the computer-executable instructions further control the one or more computing systems to identify, from the library, simulated electromagnetic data that is similar to patient electromagnetic data.
25. The one or more computing systems of claim 23 wherein a pacing lead is placed with an invasive pacing device that administers electromagnetic pulses from inside the electromagnetic source.
26. The one or more computing systems of claim 25 wherein the invasive pacing device is a catheter, an implantable cardioverter, or a pacemaker.
27. The one or more computing systems of claim 23 wherein the patient pacing is performed using a non-invasive pacing device that administers electromagnetic pulses from outside the electromagnetic source.
28. The one or more computing systems of claim 27 wherein the non-invasive pacing device generates an electromagnetic field to pace the electromagnetic source.
29. The one or more computing systems of claim 28 wherein the non-invasive pacing device is a magnetic resonance device.
30. The one or more computing systems of claim 23 wherein the electromagnetic source is a heart, the simulated electromagnetic data is a cardiogram, and the patient electromagnetic data is a cardiogram.
31. The one or more computing systems of claim 30 wherein the cardiogram is a vectorcardiogram.
32. The one or more computing systems of claim 30 wherein the cardiogram is an electrocardiogram.
33. The one or more computing systems of claim 23 wherein the configuration parameter is orientation.
34. The one or more computing systems of claim 23 wherein at least one of the identifying of simulated electromagnetic data is based on a set of previously identified simulated electromagnetic data.
35. The one or more computing systems of claim 23 wherein the computer-executable instructions further control the one or more computing systems to identify from the calibrated collection a simulated electromagnetic data that is similar to the patient electromagnetic data to identify a configuration parameter not collected from the patient.
36. The one or more computing systems of claim 35 wherein the configuration parameter is source location.
37. The one or more computing systems of claim 23 wherein the computer-executable instructions further control the one or more computing systems to identify simulated electromagnetic data that is similar to patient electromagnetic data by identifying cycles within the patient electromagnetic data and, for each of a plurality of simulated electromagnetic data of the calibrated collection, calculating a similarity score indicating similarity between the identified cycles and cycles of the simulated electromagnetic data.
38. One or more computing systems for identifying simulated electromagnetic data of an electromagnetic source that are similar to a patient electromagnetic data of a patient electromagnetic source, the simulated electromagnetic data generated based on simulated pacing locations of the electromagnetic source and the patient electromagnetic data generated based on a patient pacing locations of the patient electromagnetic source, the one or more computing systems comprising:
one or more computer-readable storage mediums for storing computer-executable instructions for controlling the one or more computing systems to:
identify simulated electromagnetic data generated based on simulated pacing locations that are similar to the patient pacing locations;
identify simulated electromagnetic data generated based on simulated action potentials that are similar to patient action potentials;
identify simulated electromagnetic data generated based on simulated conduction velocities that are similar to patient conduction velocities; and
generate training data that includes at least some of the identified simulated electromagnetic data labeled with a label that is a configuration parameter not collected from the patient; and
train a patient-specific model classifier based on the training data; and
one or more processors for executing the computer-executable instructions stored in the one or more computer-readable storage mediums.
39. The one or more computing systems of claim 38 wherein the label is a source location.
40. The one or more computing systems of claim 38 wherein the computer-executable instructions further control the one or more computing systems to apply the patient-specific model classifier to patient electromagnetic data to identify the configuration parameter that is not collected from the patient.
41. The one or more computing systems of claim 40 wherein the computer-executable instructions further control the one or more computing systems to output the identified configuration parameter to inform treatment of the patient.
42. One or more computing systems for identifying simulated electromagnetic data of an electromagnetic source that are similar to a patient electromagnetic data of a patient electromagnetic source, the simulated electromagnetic data generated based on simulated pacing locations of the electromagnetic source and the patient electromagnetic data generated based on a patient pacing locations of the patient electromagnetic source, the one or more computing systems comprising:
one or more computer-readable storage mediums for storing computer-executable instructions for controlling the one or more computing systems to:
identify simulated electromagnetic data generated based on simulated pacing locations that are similar to the patient pacing locations;
identify simulated electromagnetic data generated based on simulated action potentials that are similar to patient action potentials;
identify simulated electromagnetic data generated based on simulated conduction velocities that are similar to patient conduction velocities;
train a mapping function that maps the simulated electromagnetic data to the patient electromagnetic data; and
apply the mapping function to a set of simulated electromagnetic data to generate transformed simulated electromagnetic data wherein a calibrated collection of simulated electromagnetic data includes the transformed simulated electromagnetic data.
43. The one or more computing systems of claim 42 wherein the computer-executable instructions further control the one or more computing systems to train a patient-specific classifier based on the calibrated collection for identifying a source location based on a patient electromagnetic data.
44. One or more computing systems for identifying simulated electromagnetic data of an electromagnetic source that are similar to a patient electromagnetic data of a patient electromagnetic source, the patient electromagnetic data generated based on a patient pacing locations of the patient electromagnetic source, the one or more computing systems comprising:
one or more computer-readable storage mediums for storing computer-executable instructions for controlling the one or more computing systems to:
identify simulated electromagnetic data generated based on simulated pacing locations that are similar to the patient pacing locations, the simulated electromagnetic data derived from simulated electromagnetic output of the electromagnetic source generated by simulating, using a finite-element mesh, electrical activity of the electromagnetic source based on simulated pacing locations of the electromagnetic source;
generate training data that includes the identified simulated electromagnetic data labeled with a label that is a configuration parameter not collected from the patient; and
train a patient-specific model classifier based on the training data; and
one or more processors for executing the computer-executable instructions stored in the one or more computer-readable storage mediums.
45. The one or more computing systems of claim 44 wherein the computer-executable instructions further control the one or more computing systems to apply the patient-specific model classifier to patient electromagnetic data to identify the configuration parameter that is not collected from the patient.
46. The one or more computing systems of claim 45 wherein the computer-executable instructions further control the one or more computing systems to output the identified configuration parameter to inform treatment of the patient.
47. The one or more computing system of claim 44 wherein the configuration parameter is a source location of a fibrillation.