Calibration for continuous non-invasive blood pressure monitoring using artificial intelligence
A system for continuous non-invasive blood pressure monitoring may include processing circuitry configured to determine calibration data for a continuous non-invasive blood pressure model at a calibration point, receive, from an oxygen saturation sensing device, a PPG signal at a particular time subsequent to the calibration point, derive values of the set of metrics for the patient from the PPG signal, and determine, using the continuous non-invasive blood pressure model and based at least in part on inputting the calibration data determined at the calibration point, the values of the set of metrics, and an elapsed time at the particular time since the calibration point into the continuous non-invasive blood pressure model, a blood pressure of the patient at the particular time.
1. A method comprising:
determining calibration data for a continuous non-invasive blood pressure model at a calibration point, wherein determining the calibration data comprises:
receiving a blood pressure measurement of a patient from a blood pressure sensing device at the calibration point,
receiving a first photoplethysmographic (PPG) signal from an oxygen saturation sensing device at the calibration point, and
deriving first values of a set of metrics for the patient from the first PPG signal;
receiving a second PPG signal from the oxygen saturation sensing device at a particular time subsequent to the calibration point;
deriving second values of the set of metrics for the patient from the second PPG signal; and
determining, using the continuous non-invasive blood pressure model and based at least in part on inputting the calibration data determined at the calibration point, the second values of the set of metrics, and an elapsed time at the particular time since the calibration point into the continuous non-invasive blood pressure model, a blood pressure of the patient at the particular time.
2. The method of claim 1 , wherein the continuous non-invasive blood pressure model comprises a neural network algorithm trained via machine learning over training data that includes at least sets of calibration data at most recent calibration points of a population of patients, sets of elapsed time since the most recent calibration points, sets of PPG signals of the population of patients, and sets of target blood pressure values of the population of patients.
3. The method of claim 1 , wherein the continuous non-invasive blood pressure model comprises a neural network algorithm trained via machine learning over training data that includes at least sets of calibration data at calibration points of the patient, sets of elapsed time since the calibration points, sets of PPG signals of the patient, and sets of target blood pressure values of the patient.
4. The method of claim 1 , wherein:
the first values of the set of metrics comprises a first plurality of sequences of values over time; and
the second values of the set of metrics comprises a second plurality of sequences of values over time.
5. The method of claim 4 , wherein the set of metrics for the patient comprises one or more of: a PPG pulse duration, a PPG relative position of a maximum upslope of a systolic rise, a PPG peak location and amplitude, a PPG perfusion index, a PPG baseline trend, a PPG respiratory cycle information, a PPG upstroke area, a PPG downstroke area, a maximum gradient of a PPG upslope, or a PPG baseline value.
6. The method of claim 1 , further comprising:
periodically determining the calibration data for the continuous non-invasive blood pressure model at a plurality of calibration points by at least:
periodically receiving the blood pressure measurement for the patient from the blood pressure sensing device,
periodically receiving a PPG signal from the oxygen saturation sensing device, and
periodically deriving values of the set of metrics for the patient from the PPG signal,
wherein the blood pressure measurement of the patient at the calibration point and the first values of the set of metrics for the patient comprise the calibration data at a most recent calibration point to the particular time out of the plurality of calibration points.
7. The method of claim 6 , wherein the blood pressure measurement comprises a first blood pressure measurement, the calibration point comprises a first calibration point of the plurality of calibration points, and the particular time comprises a first particular time, the method further comprising:
determining the calibration data for the continuous non-invasive blood pressure model at a second calibration point of the plurality of calibration points by at least:
receiving a second blood pressure measurement of the patient from the blood pressure sensing device at the second calibration point,
receiving a third PPG signal from the oxygen saturation sensing device at the second calibration point, and
deriving third values of the set of metrics for the patient from the third PPG signal;
receiving a fourth PPG signal from the oxygen saturation sensing device at a second particular time subsequent to the second calibration point, wherein the second calibration point is the most recent calibration point to the second particular time;
deriving fourth values of the set of metrics for the patient from the fourth PPG signal; and
determining, using the continuous non-invasive blood pressure model and based at least in part on inputting the calibration data determined at the second calibration point, the fourth values of the set of metrics, and the elapsed time at the second particular time since the second calibration point into the continuous non-invasive blood pressure model, the blood pressure of the patient at the second particular time.
8. The method of claim 1 , wherein determining the blood pressure of the patient at the particular time further comprises:
determining, using the continuous non-invasive blood pressure model and based at least in part on the blood pressure measurements of the patient at the calibration point, the first values of the set of metrics, the second values of the set of metrics, and the elapsed time at the particular time since the calibration point, a predicted change in blood pressure between the calibration point and the particular time; and
determining the blood pressure of the patient at the particular time based at least in part on the blood pressure measurement of the patient at the calibration point and the predicted change in blood pressure between the calibration point and the particular time.
9. The method of claim 1 , wherein the calibration point comprises a specified time period.
10. A system comprising:
a blood pressure sensing device;
an oxygen saturation sensing device; and
processing circuitry configured to:
determine calibration data for a continuous non-invasive blood pressure model at a calibration point by at least:
receiving, from the blood pressure sensing device, blood pressure measurements of a patient at the calibration point,
receiving, from the oxygen saturation sensing device, a first photoplethysmographic (PPG) signal at the calibration point, and
deriving first values of a set of metrics for the patient from the first PPG signal;
receive, from the oxygen saturation sensing device, a second PPG signal at a particular time subsequent to the calibration point;
derive second values of the set of metrics for the patient from the second PPG signal; and
determine, using the continuous non-invasive blood pressure model and based at least in part on inputting the calibration data determined at the calibration point, the second values of the set of metrics, and an elapsed time at the particular time since the calibration point into the continuous non-invasive blood pressure model, a blood pressure of the patient at the particular time.
11. The system of claim 10 , wherein the continuous non-invasive blood pressure model is a neural network algorithm trained via machine learning over training data that includes at least sets of calibration data at most recent calibration points of a population of patients, sets of elapsed time since the most recent calibration points, sets of PPG signals of the population of patients, and sets of target blood pressure values of the population of patients.
12. The system of claim 10 , wherein the continuous non-invasive blood pressure model comprises a neural network algorithm trained via machine learning over training data that includes at least sets of calibration data at calibration points of the patient, sets of elapsed time since the calibration points, sets of PPG signals of the patient, and sets of target blood pressure values of the patient.
13. The system of claim 10 , wherein:
the first values of the set of metrics comprises a first plurality of sequences of values over time; and
the second values of the set of metrics comprises a second plurality of sequences of values over time.
14. The system of claim 10 , wherein the set of metrics for the patient comprises one or more of: a PPG pulse duration, a PPG relative position of a maximum upslope of a systolic rise, a PPG peak location and amplitude, a PPG perfusion index, a PPG baseline trend, a PPG respiratory cycle information, a PPG upstroke area, a PPG downstroke area, a maximum gradient of a PPG upslope, or a PPG baseline value.
15. The system of claim 10 , wherein the processing circuitry is further configured to:
periodically determine the calibration data for the continuous non-invasive blood pressure model at a plurality of calibration points by at least:
periodically receiving, from the blood pressure sensing device, the blood pressure measurement for the patient,
periodically receiving, from the oxygen saturation sensing device, a PPG signal for the patient, and
periodically deriving values of the set of metrics for the patient from the PPG signal; and
wherein the blood pressure measurement of the patient at the calibration point and the first values of the set of metrics for the patient comprise the calibration data at a most recent calibration point to the particular time out of the plurality of calibration points.
16. The system of claim 15 , wherein the blood pressure measurement comprises a first blood pressure measurement, the calibration point comprises a first calibration point out of the plurality of calibration points, the particular time comprises a first particular time, and the processing circuitry is further configured to:
determine the calibration data for the continuous non-invasive blood pressure model at a second calibration point out of the plurality of calibration points by at least:
receiving, from the blood pressure sensing device, a second blood pressure measurement of the patient at the second calibration point,
receiving, from the oxygen saturation sensing device, a third PPG signal at the second calibration point, and
deriving third values of the set of metrics for the patient from the third PPG signal;
receive, from the oxygen saturation sensing device, a fourth PPG signal at a second particular time subsequent to the second calibration point, wherein the second calibration point is the most recent calibration point to the second particular time;
derive fourth values of the set of metrics for the patient from the fourth PPG signal; and
determine, using the continuous non-invasive blood pressure model and based at least in part on inputting the calibration data determined at the second calibration point, the fourth values of the set of metrics, and the elapsed time at the second particular time since the second calibration point into the continuous non-invasive blood pressure model, the blood pressure of the patient at the second particular time.
17. The system of claim 10 , wherein the processing circuitry that is configured to determine the blood pressure of the patient at the particular time is further configured to:
determine, using the continuous non-invasive blood pressure model and based at least in part on the blood pressure measurements of the patient at the calibration point, the first values of the set of metrics, the second values of the set of metrics, and the elapsed time at the particular time since the calibration point, a predicted change in blood pressure between the calibration point and the particular time; and
determine the blood pressure of the patient at the particular time based at least in part on the blood pressure measurement of the patient at the calibration point and the predicted change in blood pressure between the calibration point and the particular time.
18. The system of claim 10 , wherein the calibration point comprises a specified time period.
19. A non-transitory computer readable storable medium comprising instructions that, when executed, cause processing circuitry to:
determine calibration data for a continuous non-invasive blood pressure model at a calibration point by at least:
receiving blood pressure measurements of a patient at the calibration point,
receiving a first photoplethysmographic (PPG) signal at the calibration point, and
deriving first values of a set of metrics for the patient from the first PPG signal;
receive a second PPG signal at a particular time subsequent to the calibration point;
derive second values of the set of metrics for the patient from the second PPG signal; and
determine, using the continuous non-invasive blood pressure model and based at least in part on inputting the calibration data determined at the calibration point, the second values of the set of metrics, and an elapsed time at the particular time since the calibration point into the continuous non-invasive blood pressure model, a blood pressure of the patient at the particular time.
20. The computer readable storable medium of claim 19 , wherein the continuous non-invasive blood pressure model is a neural network algorithm trained via machine learning over training data that includes at least sets of calibration data at most recent calibration points of a population of patients, sets of elapsed time since the most recent calibration points, sets of PPG signals of the population of patients, and sets of target blood pressure values of the population of patients.