Electrocardiogram interpretation

1 . A system for interpreting an electrocardiogram, the system comprising:

at least one processing device; and

a memory device storing instructions which, when executed by the at least one processing device, cause the at least one processing device to:

receive the electrocardiogram;

identify a segment of the electrocardiogram having an abnormality using a machine learning algorithm; and

assign an interpretation output to the segment of the electrocardiogram having the abnormality, the interpretation output describing a morphology of the abnormality, wherein the interpretation output is assigned by:

applying a range of transformations to the segment;

feeding the range of transformations through a convolutional neural network;

determining which transformation in the range of transformations produces a greatest return to normal; and

generating the interpretation output as a reverse of the transformation that produces the greatest return to normal.

2 . The system of claim 1 , wherein the interpretation output includes a textual statement.

3 . The system of claim 1 , wherein the interpretation output includes a morphology marker.

4 . The system of claim 1 , wherein the instructions, when executed by the at least one processing device, further cause the at least one processing device to:

compare the electrocardiogram to a normal electrocardiogram model;

determine deviations from the normal electrocardiogram model; and

identify the segment of the electrocardiogram having the abnormality by identifying a largest deviation from the normal electrocardiogram model.

5 . The system of claim 1 , wherein the machine learning algorithm includes saliency mapping.

6 . The system of claim 1 , wherein the machine learning algorithm includes activation mapping.

7 . A system for generating an electrocardiogram, the system including:

one or more electrodes for receiving heart electrical signals;

at least one processing device; and

a memory device storing instructions which, when executed by the at least one processing device, cause the at least one processing device to:

generate the electrocardiogram based on the heart electrical signals;

display the electrocardiogram on a display device;

identify a segment of the electrocardiogram having an abnormality using a machine learning algorithm; and

assign an interpretation output to the segment of the electrocardiogram having the abnormality, the interpretation output describing a morphology of the abnormality, wherein the interpretation output is assigned by:

applying a range of transformations to the segment;

feeding the range of transformations through a convolutional neural network;

determining which transformation in the range of transformations produces a greatest return to normal; and

generating the interpretation output as a reverse of the transformation that produces the greatest return to normal.

8 . The system of claim 7 , wherein the interpretation output includes a textual statement.

9 . The system of claim 7 , wherein the interpretation output includes a morphology marker.

10 . The system of claim 7 , wherein the instructions, when executed by the at least one processing device, further cause the at least one processing device to:

compare the electrocardiogram to a normal electrocardiogram model;

determine deviations from the normal electrocardiogram model; and

identify the segment of the electrocardiogram having the abnormality by identifying a deviation from the normal electrocardiogram model.

11 . The system of claim 7 , wherein the machine learning algorithm includes saliency mapping.

12 . The system of claim 7 , wherein the machine learning algorithm includes activation mapping.

13 . The system of claim 7 , further comprising:

one or more leads, each lead of the one or more leads having a first end for attachment to a chest of a patient, and an opposite second end operatively connected to the at least one processing device.

14 . A method of interpreting an electrocardiogram, the method comprising:

generating the electrocardiogram based on heart electrical signals;

displaying the electrocardiogram;

identifying a segment of the electrocardiogram having an abnormality using a machine learning algorithm; and

assigning an interpretation output to the segment of the electrocardiogram having the abnormality, the interpretation output describing a morphology of the abnormality, wherein the interpretation output is assigned by:

applying a range of transformations to the segment;

feeding the range of transformations through a convolutional neural network;

determining which transformation in the range of transformations produces a greatest return to normal; and

generating the interpretation output as a reverse of the transformation that produces the greatest return to normal.

15 . The method of claim 14 , further comprising:

comparing the electrocardiogram to a normal electrocardiogram model;

determining deviations from the normal electrocardiogram model; and

identifying the segment of the electrocardiogram having the abnormality by identifying a deviation from the normal electrocardiogram model.

16 . The method of claim 14 , wherein the interpretation output includes a textual statement.

17 . The method of claim 14 , wherein the machine learning algorithm includes activation mapping.