Noninvasive blood pressure measurement and monitoring

1. A sensor device, comprising:

one or more sensors configured to:

sense physiological changes associated with cardiovascular function and provide outputs corresponding to the sensed physiological changes, and

detect an electrocardiograph (ECG) signal, a photoplethysmograph (PPG) signal, and a phonocardiograph (PCG) signal of a patient from the outputs provided by the one or more sensors; and

a memory storing a program, in response to execution, causes a processor to perform operations comprising:

filtering the detected ECG, PPG, and PCG signals by detecting and rejecting artifacts from the detected ECG, PPG, and PCG signals, the artifacts including at least one presence of: arrhythmia, motion, anomalies, and outliers to produce filtered ECG, PPG, and PCG signals;

extracting multivariate signal features including magnitudes and cardiac timing intervals from the filtered ECG, PPG, and PCG signals;

averaging time series of the multivariate signal features over a predetermined time window;

processing the averaged multivariate signal features and subject-specific information using an application including an independent Blood Pressure (BP) prediction model to determine Noninvasive Blood Pressure (NIBP) including systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean blood pressure (MBP) values, wherein the independent BP prediction model is derived by machine-learning in large group individuals including demographics and disease conditions,

wherein with the independent BP prediction model, the SBP, DBP, and MBP values of the NIBP are mapped according to the averaged multivariate signal features and the subject-specific information; and

displaying or continuously logging the SBP, DBP, and MBP values.

2. The sensor device of claim 1 , wherein the processor performs further operations comprising:

displaying or continuously logging pulse pressure; wherein the displaying or continuously logging is at a rate including beat, second, minute, and/or hour.

3. The sensor device of claim 1 , wherein the processor performs further operations comprising displaying or continuously logging an indication of the NIBP exceeding a threshold.

4. The sensor device of claim 1 ,

wherein the processor performs further operations comprising: outputting the multivariate signal features formatted for one or more of machine or human readability;

wherein the outputted multivariate signal features include the multivariate signal features derived from the detected ECG, PPG and/or PCG signals, including one or more of cardiac timing intervals or core perfusion intensity; and

wherein the outputted multivariate signal features are one or more of predictors or correlates of the NIBP.

5. The sensor device of claim 1 , wherein the determined NIBP measurement is calibration-free.

6. The sensor device of claim 1 , wherein the processor performs further operations comprising displaying or continuously logging an indicator of blood pressure other than a numeric blood pressure value.

7. The sensor device of claim 1 , further comprising a base configured to physically support the one or more sensors adapted to be worn by a subject of the sensed physiological changes associated with cardiovascular function.

8. The sensor device of claim 6 , wherein the indicator is a non-numeric output.

9. A method to determine noninvasive blood pressure, comprising:

sensing, by a sensor, physiological changes associated with cardiovascular function;

detecting, by the sensor, electrocardiograph (ECG), photoplethysmograph (PPG), and phonocardiograph (PCG) signals based on the sensed physiological changes;

filtering, by a processor, the detected ECG, PPG, and PCG signals by detecting and rejecting artifacts from the detected ECG, PPG, and PCG signals, the artifacts including at least one presence of: arrhythmia, motion, anomalies, and outliers to produce filtered ECG, PPG, and PCG signals;

extracting, by the processor, multivariate signal features including magnitudes and cardiac timing intervals from the filtered ECG, PPG, and PCG signals;

averaging time series of the multivariate signal features over a predetermined time window;

processing, by the processor, the averaged multivariate signal features and subject-specific information using an application including an independent Blood Pressure (BP) prediction model to determine Noninvasive Blood Pressure (NIBP) including systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean blood pressure (MBP) values, wherein the independent BP prediction model is derived by machine-learning in large group individuals including demographics and disease conditions,

wherein with the independent BP prediction model, the SBP, DBP, and MBP values of the NIBP are mapped according to the averaged multivariate signal features and the subject-specific information; and

displaying or continuously logging the SBP, DBP, and MBP values.

10. The method of claim 9 , further comprising:

displaying or continuously logging pulse pressure; and

wherein the displaying or continuously logging is at a rate including beat, second, minute, and/or hour.

11. The method of claim 9 , further comprising displaying or continuously logging an indication of the NIBP exceeding a threshold.

12. The method of claim 9 , further comprising:

outputting, by the processor, the multivariate signal features formatted for one or more of machine or human readability;

wherein the outputted multivariate signal features include the multivariate signal features derived from the detected ECG, PPG and/or PCG signals, including one or more of cardiac timing intervals or core perfusion intensity; and

wherein the outputted multivariate signal features are one or more of predictors or correlates of the NIBP.

13. The method of claim 12 , wherein the determined NIBP measurement is calibration-free.

14. The method of claim 9 , further comprising displaying or continuously logging an indicator of the blood pressure condition other than a numeric blood pressure value.

15. The method of claim 14 , wherein the indicator is a non-numeric output.

16. The method of claim 9 , further comprising:

attaching, to a subject of the sensing, a base configured to physically support the sensor;

wherein the sensor performs the sensing of the physiological changes associated with cardiovascular function of the subject.

17. A system to predict Noninvasive Blood Pressure (NIBP), comprising:

a sensor device including:

one or more sensors configured to:

sense physiological changes associated with cardiovascular function and provide outputs corresponding to the sensed physiological changes, and

detect an electrocardiograph (ECG) signal, a photoplethysmograph (PPG) signal, and a phonocardiograph (PCG) signal from the outputs provided by the one or more sensors;

a processor configured to:

filter the detected ECG, PPG, and PCG signals by detecting and rejecting artifacts from the detected ECG, PPG, and PCG signals, the artifacts including at least one presence of: arrhythmia, motion, anomalies, and outliers to produce filtered ECG, PPG, and PCG signals;

extract multivariate signal features including magnitudes and cardiac timing intervals from the filtered ECG signal, PPG signal, and PCG signal;

average time series of the multivariate signal features over a predetermined time window;

and

a wireless transmitter configured to transmit wirelessly the multivariate signal features to a computer; and

the computer configured to:

process the averaged multivariate signal features and subject-specific information using an application including an independent Blood Pressure (BP) prediction model to determine Noninvasive Blood Pressure (NIBP) including systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean blood pressure (MBP) values, wherein the independent BP prediction model is derived by machine-learning in large group individuals including demographics and disease conditions,

wherein with the independent BP prediction model, the SBP, DBP, and MBP values of the NIBP are mapped according to the averaged multivariate signal features and the subject-specific information; and

display or continuously log the SBP, DBP, and MBP values.

18. The system of claim 17 , wherein the computer is further configured to display or continuously log an indication of the NIBP exceeding a threshold.

19. The system of claim 17 ,

wherein the computer is further configured to output the multivariate signal features formatted for one or more of machine or human readability;

wherein the outputted multivariate signal features include the multivariate signal features derived from the detected ECG, PPG and/or PCG signals, including one or more of cardiac timing intervals or core perfusion intensity; and

wherein the outputted multivariate signal features are one or more of predictors or correlates of the NIBP.

20. The system of claim 19 ,

wherein the determined NIBP measurement is calibration-free.