Noninvasive methods for detection of pulmonary hypertension
Inventors
Wagner, Tyler • AWASTHI, Samir • Soundararajan, Venkataramanan • Aravamudan, Murali • CARPENTER, Corinne • Carlson, Katherine • Attia, Itzhak Zachi • Friedman, Paul A. • Asirvatham, Samuel J. • Kapa, Suraj • Lopez-Jimenez, Francisco • Dubrock, Hilary M.
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12340906-B2
Publication Date
2025-06-24
Expiration Date
Abstract
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.
Core Innovation
The invention relates to noninvasive pulmonary hypertension detection using electrocardiogram voltage-time data of a subject. The voltage-time data comprises an electrocardiogram waveform of a plurality of leads of an electrocardiograph, and the method generates a plurality of feature vectors from portions of the ECG waveform and provides the feature vectors to a pretrained learning system that generates fixed size encodings and predictions.
The pretrained learning system receives the plurality of feature vectors and outputs predictions that are used to receive an indication of presence or absence of pulmonary hypertension in the subject. The indication is then provided to a computing node for display to a user, and the method and system support providing the indication to an electronic health record system for storage in a health record associated with the subject.
The invention further includes feature-vector generation options and pretrained learning system architecture details. The generating of feature vectors can comprise producing at least one spectrogram from voltage time data for multiple leads and grouping the voltage data from the multiple leads into multiple subsets. The pretrained learning system can comprise a machine-learning model configured as a convolutional neural network with at least one residual connection.
Claims Coverage
The independent claims are directed to a method, a system, and a computer program product. Across these claims, the core inventive features include generating feature vectors from ECG voltage-time data, using a pretrained learning system to produce fixed size encodings and predictions, and outputting an indication of presence or absence of pulmonary hypertension for display to a user.
Receiving voltage-time ECG waveform data and generating feature vectors from waveform portions
Receiving voltage-time data of a subject, the voltage-time data comprising an electrocardiogram waveform of a plurality of leads of an electrocardiograph, and generating a plurality of feature vectors from the voltage-time data, wherein each of the feature vectors comprises a portion of the ECG waveform.
Pretrained learning system producing fixed size encodings and predictions
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.
Outputting an indication of presence or absence of pulmonary hypertension for user display
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.
System including electrocardiograph and computing node to perform pulmonary hypertension detection
Providing a system comprising an electrocardiograph comprising a plurality of leads and a computing node operating a processor, the computing node including program instructions executable to receive voltage-time data of a subject from voltage-time data comprising ECG waveform of the plurality of leads, generate a plurality of feature vectors from the voltage-time data, provide the plurality of feature vectors to a pretrained learning system, receive an indication of presence or absence of pulmonary hypertension in the subject based on predictions, and provide the indication to a computing node for display to a user.
Computer program product executing pulmonary hypertension detection method
Providing a computer program product comprising a non-transitory computer readable storage medium having program instructions executable by a processor to cause the processor to receive voltage-time data of a subject comprising ECG waveform of the plurality of leads, generate a plurality of feature vectors from the voltage-time data, provide the plurality of feature vectors to a pretrained learning system to generate fixed size encodings and predictions, receive an indication of a presence or absence of pulmonary hypertension in the subject based on the predictions, and provide the indication to a computing node for display to a user.
The inventive coverage focuses on converting ECG voltage-time waveform data from a plurality of leads into a plurality of feature vectors, using a pretrained learning system to generate fixed size encodings and predictions, and producing an indication of presence or absence of pulmonary hypertension that is provided for user display. The claims also include feature generation options, pretrained learning system architecture, and output routing to an electronic health record system.
Stated Advantages
Not explicitly described in patent.
Documented Applications
Not explicitly described in patent.
Interested in licensing this patent?