Noninvasive methods for detection of pulmonary hypertension
Provided herein are methods, systems, and computer program products for the detection of pulmonary hypertension comprising receiving voltage-time data of a plurality of leads of an electrocardiograph of a subject; generating a feature vector from the voltage-time data; providing the feature vector to a pretrained learning system; and receiving from the pretrained learning system an indication of the presence or absence of pulmonary hypertension in the subject.
1. A method comprising:
receiving voltage-time data of a subject, the voltage-time data comprising an electrocardiogram (ECG) waveform of a plurality of leads of an electrocardiograph;
generating a plurality of feature vectors from the voltage-time data, wherein each of the feature vectors comprises a portion of the ECG waveform;
providing the plurality of the feature vectors to a pretrained learning system, wherein the pretrained learning system comprises:
a machine-learning model configured to:
generate a plurality of fixed size encodings from the plurality of feature vectors respectively; and
generate predictions based on the plurality of fixed size encodings;
receiving from the pretrained learning system an indication of a presence or absence of pulmonary hypertension in the subject based on the predictions; and
providing the indication to a computing node for display to a user.
2. The method of claim 1 , wherein generating the feature vectors comprises at least one of generating a spectrogram based on voltage data of the plurality of leads, and grouping the voltage data of the plurality of leads into a plurality of subsets.
3. The method of claim 1 , further comprising at least one of:
receiving demographic information of the subject, wherein generating the feature vectors comprises adding the demographic information to the feature vectors; and
receiving genomic information of the subject, wherein generating the feature vectors comprises adding the genomic information to the feature vectors.
4. The method of claim 1 , wherein the pretrained learning system comprises a convolutional neural network that comprises at least one residual connection.
5. The method of claim 1 , wherein the voltage-time data of a subject is received from at least one of an electrocardiograph and an electronic medical record.
6. The method of claim 1 , further comprising: providing the indication to an electronic health record system for storage in a health record associated with the subject.
7. A system comprising:
an electrocardiograph comprising a plurality of leads;
a computing node, operating a on a processor, comprising a non-transitory computer readable storage medium having program instructions embodied therewith, the program instructions executable by the processor of the computing node to cause the processor to perform a method comprising:
receiving voltage-time data of a subject from voltage-time data comprising electrocardiogram (ECG) waveform of the plurality of leads;
generating a plurality of feature vectors from 8the voltage-time data, wherein each of the feature vectors comprises a portion of the ECG waveform;
providing the plurality of feature vectors to a pretrained learning system, wherein the pretrained learning system comprises:
a machine learning model configured to:
generate a plurality of fixed size encodings from the plurality of feature vectors respectively; and
generate predictions based on the plurality of fixed size encodings;
receiving from the pretrained learning system an indication of presence or absence of pulmonary hypertension in the subject based on the predictions; and
providing the indication to a computing node for display to a user.
8. The system of claim 7 , wherein generating the feature vectors comprises at least one of generating a spectrogram based on the voltage time data of the plurality of leads, and grouping the voltage data of the plurality of leads into a plurality of subsets.
9. The system of claim 7 , further comprising at least one of:
receiving demographic information of the subject, wherein generating the feature vectors comprises adding the demographic information to the feature vectors, and
receiving genomic information of the subject, wherein generating the feature vectors comprises adding the genomic information to the feature vectors.
10. The system of claim 7 , wherein the pretrained learning system comprises a convolutional neural network that comprises at least one residual connection.
11. The system of claim 7 , wherein the voltage-time data of a subject is received from at least one of an electrocardiograph and an electronic medical record.
12. The system of claim 7 , further comprising providing the indication to an electronic health record system for storage in a health record associated with the subject.
13. A computer program product for detection of pulmonary hypertension, the computer program product comprising a non-transitory computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to perform a method comprising:
receiving voltage-time data of a subject from voltage-time data comprising electrocardiogram (ECG) waveform of a plurality of leads;
generating a plurality of feature vectors from the voltage-time data, wherein each of the feature vectors comprises a portion of the ECG waveform;
providing the plurality of feature vectors to a pretrained learning system, wherein the pretrained learning system comprises:
a machine learning model configured to:
generate a plurality of fixed size encodings from the plurality of feature vectors respectively; and generate predictions based on the plurality of fixed size encodings;
receiving from the pretrained learning system an indication of a presence or absence of pulmonary hypertension in the subject based on the predictions; and
providing the indication to a computing node for display to a user.
14. The computer program product of claim 13 , wherein generating the feature vectors comprises at least one of generating a spectrogram based on the voltage-time data of the plurality of leads, and grouping the voltage data of the plurality of leads into a plurality of subsets.
15. The computer program product of claim 13 , further comprising at least one of:
receiving demographic information of the subject, wherein generating the feature vectors comprises adding the demographic information to the feature vectors; and
receiving genomic information of the subject, wherein generating the feature vectors comprises adding the genomic information to the feature vectors.
16. The computer program product of claim 13 , wherein the pretrained learning system comprises a convolutional neural network that comprises at least one residual connection.
17. The computer program product of claim 13 , wherein the voltage-time data of a subject is received from at least one of an electrocardiograph and an electronic medical record.
18. The computer program product of claim 13 , further comprising:
providing the indication to an electronic health record system for storage in a health record associated with the subject.