Two-lead QT interval prediction
Embodiments of the present disclosure provide a mobile electrocardiogram (ECG) sensor comprising an electrode assembly comprising electrodes, wherein the electrode assembly senses heart-related signals when in contact with a body of a user, and produces electrical signals representing the sensed heart-related signals. The ECG sensor further comprises a processing device, operatively coupled to the electrode assembly, the processing device to provide the sensed heart-related signals to a machine learning module trained to predict a twelve-lead QT interval (QTc) value from the mobile ECG sensor comprising less than twelve leads. The ECG sensor also comprises a housing containing the electrode assembly and the processing device.
1. A mobile electrocardiogram (ECG) sensor comprising:
an electrode assembly comprising electrodes, wherein the electrode assembly is configured to sense heart-related signals when in contact with a body of a user, and produce electrical signals representing the sensed heart-related signals;
a processing device, operatively coupled to the electrode assembly, the processing device to:
provide a training data set comprising a plurality of electrocardiogram (ECG) measurements and a corresponding QT interval label for each of the plurality of ECG measurements;
train a machine learning (ML) model by:
analyzing, using the ML model, each of the plurality of ECG measurements to generate an output;
comparing each of the plurality of outputs to a corresponding QT interval label to generate an error that is based on a cross entropy term of the output and the corresponding QT interval label; and
updating the ML model based on the error generated for each of the plurality of outputs;
analyze the sensed heart-related signals using the machine learning model to predict a twelve-lead QT interval (QTc) value based on the sensed heart-related signals, wherein the sensed heart-related signals comprise less than twelve leads; and
analyze, using the machine learning model, the predicted QTc value to determine whether a health anomaly is present; and
a housing containing the electrode assembly and the processing device.
2. The mobile ECG sensor of claim 1 , wherein the plurality of electrocardiogram (ECG) measurements comprise twelve-lead ECG measurements from a plurality of users and the ML model is trained to predict the twelve-lead QTc value for a single user.
3. The mobile ECG sensor of claim 1 , wherein the plurality of electrocardiogram (ECG) measurements comprise twelve-lead ECG measurements from a single user and the ML model is trained to predict the twelve-lead QTc value for the single user.
4. The mobile ECG sensor of claim 1 , wherein the ML model is a deep neural network machine learning module.
5. The mobile ECG sensor of claim 1 , wherein the sensed heart-related signals comprise Lead I and Lead II signals.
6. The mobile ECG sensor of claim 1 , wherein the health anomaly is determined to be present when QTc prolongation is detected.
7. The mobile ECG sensor of claim 1 , wherein the processing device is further configured to send a notification to a client device in response to the health anomaly being present.
8. The mobile ECG sensor of claim 6 , wherein QTc prolongation corresponds to an increase in QTc that is above a threshold amount and within a threshold amount of time.
9. The mobile ECG sensor of claim 1 , wherein the processing device determines that the health anomaly is present when QTc prolongation is detected from the predicted QTc value.
10. The mobile ECG sensor of claim 1 , wherein the processing device is further to send a notification to a client device in response to determining that the health anomaly is present.
11. A mobile electrocardiogram (ECG) system comprising:
an electrode assembly comprising electrodes, wherein the electrode assembly is configured to sense heart-related signals when in contact with a body of a user, and produce electrical signals representing the sensed heart-related signals;
a processing device, operatively coupled to the electrode assembly, the processing device to:
provide a training data set comprising a plurality of electrocardiogram (ECG) measurements and a corresponding QT interval label for each of the plurality of ECG measurements;
train a machine learning (ML) model by:
analyzing, using the ML model, each of the plurality of ECG measurements to generate an output;
comparing each of the plurality of outputs to a corresponding QT interval label to generate an error that is based on a cross entropy term of the output and the corresponding QT interval label; and
updating the ML model based on the error generated for each of the plurality of outputs;
analyze the sensed heart-related signals using the ML model to predict a twelve-lead QT interval (QTc) value based on the sensed heart-related signals, wherein the sensed heart-related signals comprise less than twelve leads; and
analyze, using the ML model, the predicted QTc value to determine whether a health anomaly is present;
a display operably connected to the processing device; and
a memory comprising instructions to cause the processing device to process the sensed heart-related signals and display the heart-related signals and the predicted QTc value on the display.
12. The mobile ECG system of claim 11 , wherein the ML model is trained on twelve-lead QTc interval data from a plurality of users to predict the twelve-lead QTc value for a single user.
13. The mobile ECG system of claim 11 , wherein the ML model is trained on twelve-lead QTc interval data from a single user to predict the twelve-lead QTc value for the single user.
14. The mobile ECG system of claim 11 , wherein the ML model is a deep neural network machine learning module.
15. The mobile ECG system of claim 11 , wherein the sensed heart-related signals comprise Lead I and Lead II signals.
16. The mobile ECG sensor of claim 11 , wherein the health anomaly is determined to be present when QTc prolongation is detected.
17. The mobile ECG sensor of claim 11 , wherein the processing device is further configured to send a notification to a client device in response to the health anomaly being present.
18. The mobile ECG system of claim 16 , wherein QTc prolongation corresponds to an increase in QTc that is above a threshold amount and within a threshold amount of time.
19. The mobile ECG system of claim 11 , wherein the processing device determines that the health anomaly is present when QTc prolongation is detected from the predicted QTc value.
20. The mobile ECG sensor of claim 11 , wherein the processing device is further to send a notification to a client device in response to determining that the health anomaly is present.
21. A method, comprising:
providing a training data set comprising a plurality of electrocardiogram (ECG) measurements and a corresponding QT interval label for each of the plurality of ECG measurements;
training a machine learning (ML) model by:
analyzing, using the ML model, each of the plurality of ECG measurements to generate an output;
comparing each of the plurality of outputs to a corresponding QT interval label to generate an error that is based on a cross entropy term of the output and the corresponding QT interval label; and
updating the ML model based on the error generated for each of the plurality of outputs;
receiving, from an electrode assembly comprising electrodes, heart-related signals sensed by the electrode assembly from a body of a user;
generating electrical signals representing the sensed heart-related signals;
analyzing, by a processing device, the heart-related signals using the ML model to predict a twelve-lead QT interval (QTc) value based on the sensed heart-related signals, wherein the sensed heart-related signals comprise less than twelve leads; and
analyzing, by the processing device, the predicted QTc value using the ML model to determine whether a health anomaly is present.
22. The method of claim 21 , wherein the plurality of electrocardiogram (ECG) measurements comprise twelve-lead ECG measurements from a plurality of users to predict the twelve-lead QTc value for a single user.
23. The method of claim 21 , further comprising determining the health anomaly to be present when QTc prolongation is detected from the predicted QTc value.
24. The method of claim 21 , further comprising sending a notification to a client device in response to the health anomaly being present.
25. The method of claim 23 , wherein QTc prolongation corresponds to an increase in QTc that is above a threshold amount and within a threshold amount of time.